Abstract
Purpose:
Although return to learn, exercise, and sports have evidence-based guidelines, there is limited research investigating return to driving after concussion. The purpose was to characterize and compare adolescent driving behaviors after concussion.
Methods:
Using the Minds Matter Concussion Registry, we queried data of adolescents, ages 16-19, diagnosed with a concussion ≤28 days of injury and seen 1/31/17-8/31/18 at the specialty care concussion program. Outcomes included patient report of: changes post-injury driving behaviors; Post-Concussion Symptom Inventory (PCSI); return to school, and exercise and sports. Provider recommendations for return to school after initial clinical assessment were also examined. Descriptive statistics, ANCOVA, and chi-square tests were performed.
Results:
Of the 332 drivers (46.1% female; mean age 17.5 years, 95% CI, 17.4-17.6), 46.9% had returned to driving since injury. Of those who returned to driving, 58.9% reported “Driving with No Changes.” PCSI score was higher among “Driving with Changes” (48.7, 95% CI: 42.2-55.2) than “Driving with No Changes” (27.4, 95% CI: 22.3-32.5, p<0.01) and “Has Not Driven Since Injury” (42.3, 95% CI: 38.4-46.3, p<0.01). Among the 332 drivers, few had returned to exercise (15.4%) or organized sports (6.0%). Of those in school (n=291), only 8.9% were provider recommended to return to full school days after clinical assessment.
Conclusion:
Many adolescents continued to drive after concussion, despite not yet having returned to exercise or sport. Nine out of ten were advised to return to school with accommodations to begin a gradual increase in cognitive activity, suggesting a gradual increase in driving may be justified.
Introduction
Each year, over 1.9 million children sustain a concussion, with adolescents accounting for over 50% of these mild traumatic brain injuries (mTBI).[1] A concussion can affect cognition, concentration, processing speed, and oculomotor function.[2–5] Driving a vehicle is cognitively demanding, requiring integration of these functions in rapid succession, including visual scene assessment, processing of environmental risks, and execution of complex skills. Therefore, decrements due to concussion can impair abilities integral to safe driving. Even though there are guidelines on return to learn, exercise and sports, there are no evidence-based guidelines for return to driving post-injury for adolescents. The general consensus is that acutely injured, symptomatic adolescents should not drive;[6–8] however after the symptomatic acute phase, there is a lack of evidence to guide clinical recommendations.[9,10] This is of particular concern for adolescents because adolescent drivers have the highest motor vehicle crash rates among drivers [11–13] and account for a substantial percentage of concussions.
Little research has examined return to drive in adolescents after concussion. The few studies examining return to driving after concussion have focused primarily on adult drivers, or combined adult and adolescent samples.[14,15] Studies have shown that self-report symptom burden can influence return to drive, including impairing one’s ability to judge readiness to drive, concentration, and memory.[2,3,16] In a simulator study examining driving performance of adolescents and young adults, concussed participants, even those who were asymptomatic, had more unsafe driving behaviors, such as lane excursions, mean standard deviation of lane position, and greater standard deviation of speed maintenance.[17] In addition, even though adolescents and young adults believe driving immediately after a concussion is unsafe, they self-report return to drive soon after a concussion,[18] with a lack of standardized guidance from their healthcare providers.[7–9] Since the first two years of licensure are characterized by rapid skill development[19] and adolescents have different driving practices and crash risks than young adults,[13] the intersection of driving and concussion recovery needs to be examined uniquely in adolescents as a particularly high-risk population. Therefore, the objective of this study was to characterize adolescent driving behaviors after concussion and examine clinical differences across groups stratified by their post-concussion driving practices.
Patients and Methods
Data for this analysis were queried from the Minds Matter Concussion Registry, which prospectively collects data from the electronic health record (EHR) for patients seen for concussion across the care network at the Children’s Hospital of Philadelphia. For this cross-sectional analysis, we queried data for adolescent patients ages 16-19 years diagnosed with a concussion ≤28 days of injury and seen between 1/31/17-8/31/18 for their initial clinical visit to the specialty care concussion program for their current injury. Patients were diagnosed with a concussion according to the definition put forth in the Consensus Statement on Concussion in Sport, including history of an injury with transmitted forces to the head followed by the onset of symptoms associated with concussion.[20]
As part of their visit to the specialty care concussion program and the Minds Matter Concussion Registry, adolescent patients completed a questionnaire. For this analysis, responses from the item identifying pre- to post-injury changes in their driving were used. The item asked: “Has the patient changed how he/she drives since the injury?” The response options included: “No;” “Not a driver;” “Has not driven since injury;” “Limits number of trips;” “Limits distance of trips;” “Increases space between vehicles;” “Avoids certain traffic conditions or geographic locations;” and “Avoids driving at night.” Patients could choose multiple categories of changes made to driving practices. For this analysis, participants were categorized as “Not a Driver,” “Has Not Driven Since Injury,” “Driving with No Changes,” or “Driving with Changes.” “Driving with Changes” included the responses of yes to one or more of the categories of limiting trips, limiting trip distance, increasing space between vehicles, avoiding certain condition and locations and avoids driving at night. Those who identified as “Not a Driver” were not further analyzed.
As part of the questionnaire, patients completed the Post-Concussion Symptom Inventory (PCSI),[21] self-report of symptom severity over the past two days (21 items on a Likert scale of 0-6 (0=none, 6=most severe), range of total score 0-126). The PCSI includes subscales of physical, fatigue, emotional and cognitive domains. The PCSI has been shown to have good internal consistency (α=0.94), good test-re-test (ICC=0.79) and correlation with parent-report (r=0.65).[21] PCSI total and subscale score were used in analysis.
Results of the questionnaire for pre- to post-injury changes in driving and the PCSI were accessed via a data export from the clinical registry, as well as data on the following: age, sex, race, ethnicity, mechanism of injury, days since injury, prior history of physician-diagnosed concussion, and information on self-reported status of return to school, exercise, and organized sports since the injury (i.e. their status prior to initial visit’s clinical assessment). Lastly, the clinician’s recommendation for return to school after initial clinical assessment was also accessed via a data export from the EHR. The clinician’s recommendation for return to school contained details about recommended school accommodations such as partial versus full day, breaks as needed during the day, reduced workload, and extra time for assignments and testing. Any missing data in the export were obtained via supplemental medical record review, if available.
Descriptive statistics were used to describe the participant demographics and characterize changes to driving behaviors. A series of analysis of variance (ANOVAs), chi-square analyses, and Fisher’s exact tests were performed to examine differences among the three groups of driving behavior changes pre-to post-injury: “Driving with No Changes,” “Driving with Changes,” and “Has Not Driven Since Injury.” Post-hoc testing for bivariate group comparisons were done with Tukey’s range test, chi-square analyses, and Fisher’s exact tests as appropriate. Age, days since injury and sex were used as covariates in group comparisons using analysis of covariance ANCOVA for continuous variables. Analyses were conducted using R. This study was approved by the Institutional Review Board at the Children’s Hospital of Philadelphia and had a waiver of consent for use of the data collected in the EHR for the Minds Matter Concussion Registry.
Results
A total of 491 adolescents met the inclusion criteria. One hundred and fifty-three reported that they were “Not a Driver” and were not included in analysis. Six patients were further excluded as they did not complete the questionnaire item identifying driving practices post-injury. Table 1 includes the demographic characteristics of the 332 adolescents who reported being drivers. The sample was mostly male (53.9%), non-Hispanic white (83.4%), and on average 17.5 years old (95% CI: 17.4-17.6). Adolescents presented to specialty care concussion program an average 12.3 (95% CI: 11.6-13.1) days post-injury (range 1-28 days). The most common mechanism of injury was struck by/against object (n=132), followed by struck by person (n=78). There were significant differences in demographics or injury related variables between those who did not complete the item on driving practices post-injury (n=6) and those included in analysis (n=332). However, those who were not a driver (n=153) were significantly younger than the drivers (n=332) (16.9, 95% CI: 16.8-17.0 v. 17.5, 95% CI: 17.4-17.6, p < 0.001), further away from injury (13.8 days, 95% CI: 12.7-14.9 v. 12.3 days, 95% CI: 11.6-13.1, p = 0.022) and were less likely to be NH White (54.2% NH white v 83.4% NH white, p < 0.001). No other differences were found in demographics or injury related variables between the drivers and those not included in further analyses.
Table 1.
Patient Characteristics by Group
| Total Sample (N = 332) | Driving with No Changes (N = 92) | Driving with Changes (N = 64) | Has Not Driven Since Injury (N = 176) | p-value | |
|---|---|---|---|---|---|
| Age, mean (95% CI), years | 17.5 (17.4, 17.6) | 17.6 (17.4, 17.7) | 17.7 (17.5, 17.9) | 17.4 (17.2, 17.5) | 0.03 |
| Sex, No. (%) Female | 153 (46.1%) | 51 (55.4%) | 30 (46.9%) | 72 (40.9%) | 0.08 |
| Race/ethnicity, No. (%)a | 0.16 | ||||
| Non-Hispanic white | 277 (83.4%) | 82 (89.1%) | 54 (84.4%) | 141 (80.1%) | |
| Non-Hispanic black | 15 (4.5%) | 4 (4.3%) | 0 (0.0%) | 11 (6.2%) | |
| Hispanic | 13 (3.9%) | 1 (1.1%) | 4 (6.2%) | 8 (4.5%) | |
| Non-Hispanic Asian/Asian Pacific Islander/other/multiple race/unknown | 27 (8.1%) | 5 (5.4%) | 6 (9.4%) | 16 (9.1%) | |
| Days since injury, mean (95% CI) | 12.3 (11.6, 13.1) | 12.9 (11.4, 14.4) | 13.9 (12.3, 15.5) | 11.5 (10.4, 12.5) | 0.04 |
| Prior history of concussion before current injury, No. (%)b | 178 (53.8%) | 57 (62.0%) | 35 (54.7%) | 86 (49.1%)b | 0.13 |
| Mechanism of injury, No. (%) | 0.57 | ||||
| Assault | 6 (1.8%) | 2 (2.2%) | 1 (1.6%) | 3 (1.7%) | |
| Fall | 68 (20.5%) | 18 (19.6%) | 7 (10.9%) | 43 (24.4%) | |
| Pedestrian struck | 1 (0.3%) | 0 (0.0%) | 0 (0.0%) | 1 (0.6%) | |
| Passenger in car | 23 (6.9%) | 5 (5.4%) | 3 (4.7%) | 15 (8.5%) | |
| Struck by person | 78 (23.5%) | 24 (26.1%) | 18 (28.1%) | 36 (20.5%) | |
| Struck by/against object | 132 (39.8%) | 34 (37.0%) | 30 (46.9%) | 68 (38.6%) | |
| MVA | 9 (2.7%) | 2 (2.2%) | 2 (3.1%) | 5 (2.8%) | |
| Other | 15 (4.5%) | 7 (7.6%) | 3 (4.7%) | 5 (2.8%) |
Chi-square analysis done with NH White vs all other categories grouped as Other.
One participant with missing data.
Of the 332 adolescent drivers, 46.9% had returned to driving since injury (n=156). Of those who returned to drive, 58.9% (n=92) reported “Driving with No Changes.” There were no differences in sex, race/ethnicity, history of prior concussion or mechanism of injury among the three driving groups (e.g. “Driving with No Changes” (n=92), “Driving with Changes” (n=64), and “Has Not Driven Since Injury” (n=176)). However, there were statistically significant differences in age (p=0.03) and days since injury (p=0.04) across groups. Post-hoc testing revealed that those reporting “Has Not Driven Since Injury” were slightly younger (17.4, 95% CI: 17.2-17.5) than those “Driving with Changes” (17.7, 95% CI: 17.5-17.9) (p = 0.026). Post-hoc testing also revealed that those reporting “Has Not Driven Since Injury” were approximately 2 days closer to injury (11.5, 95% CI: 10.4, 12.5) than those “Driving with Changes” (13.9, 95% CI: 12.3, 15.5) (p = 0.035). Age and days since injury were thus included as covariates subsequent analyses that used ANCOVAs. While sex was not statistically different between groups, it was also included as a covariate in ANCOVAs as prior studies have shown sex differences in self-reported symptom severity on PCSI[22] and in driving behaviors.[23]
Of the 64 adolescents who made changes to their driving behaviors, limiting the number of trips was the most common change (n=45), followed by limiting the distance of trips (n=15), and avoiding driving at night (n=14) (Table 2). Forty-six adolescents made only one change, which predominantly was limiting the number of trips (n=31). Eleven adolescents made two changes, three made three changes, three made four changes, and one made five changes.
Table 2.
Number Making Each Change and Total Number and Combinations for One or More Changes
| Types of Change No. (%) |
One Change (N = 46) | Two Changes (N = 11) | Three Changes (N = 3) | Four Changes (N = 3) | Five Changes (N = 1) |
|---|---|---|---|---|---|
| Limits number of trips (n=45) | 31 (67.4%) | 8 (72.7%) | 3 (100%) | 3 (100%) | 0 (0%) |
| Limits distance of trips, (n=15) | 4 (8.7%) | 6 (54.5%) | 3 (100%) | 1 (33.3%) | 1 (100%) |
| Increase space (n=8) | 2 (4.3%) | 2 (18.2%) | 1 (33.3%) | 2 (66.7%) | 1 (100%) |
| Avoids locations (n=8) | 2 (4.3%) | 4 (36.4%) | 0 (0%) | 1 (33.3%) | 1 (100%) |
| Limits lane changes (n=4) | 1 (2.2%) | 0 (0%) | 0 (0%) | 2 (66.7%) | 1 (100%) |
| Avoids driving at night (n=14) | 6 (13%) | 2 (18.2%) | 2 (66.7%) | 3 (100%) | 1 (100%) |
Table 3 displays self-report of concussion symptoms (PCSI) and return to activity. The range of score of PCSI in the sample was 0-105. When examining PCSI among the three groups of post-injury driving behavior, total PCSI score differed among groups (p<0.01), with post-hoc testing revealing that adolescents who were “Driving with No Changes” had a significantly lower total PCSI score (27.4, 95% CI: 22.3-32.5) than adolescents who reported “Driving with Changes” (48.7, 95% CI: 42.2-55.2, p<0.01) and “Has Not Driven Since Injury” (42.3, 95% CI: 38.4-46.3, p<0.01). This pattern of significant differences extended to the physical, fatigue and cognitive subscales of the PCSI (Table 3). For the emotional subscale, those that were “Driving with Changes” had higher scores than the “Driving with No Changes” and “Has Not Driven Since Injury.” When examining the proportion of adolescents in lower quartile of PCSI scores (<15), those who were “Driving with No Changes” had a higher proportion with scores < 15 (n=40, 43.5%) compared to “Driving with Changes” (n=8, 12.5%) and “Has Not Driven Since Injury” (n=33, 19%) (p< 0.01)
Table 3.
Self-report Concussion Symptoms and Return to Activity
| Total Samplea (N = 332) | Driving with No Changes (N = 92) (0) | Driving with Changes (N = 64) (1) | Has Not Driven Since Injurya (N = 176) (2) | p-value | Post-hoc | |
|---|---|---|---|---|---|---|
| PCSI total, mean (95% CI) | 39.4 (36.5, 42.3) | 27.4 (22.3, 32.5) | 48.7 (42.2, 55.2) | 42.3 (38.4, 46.3) | <0.01 | 1-0: <0.01 2-0: <0.01 2-1: 0.275 |
| PCSI physical | 14.2(13.0, 15.3) | 9.4 (7.6, 11.3) | 17.1 (14.7, 19.5) | 15.6 (14.0, 17.1) | <0.01 | 1-0: <0.01 2-0: <0.01 2-1: 0.49 |
| PCSI fatigue | 7.3 (6.8, 7.9) | 5.3 (4.2, 6.4) | 8.5 (7.3, 9.7) | 8.0 (7.3, 8.7) | <0.01 | 1-0: <0.01 2-0: <0.01 2-1: 0.81 |
| PCSI emotional | 5.3 (4.6, 5.9) | 3.7 (2.7, 4.7) | 7.6 (6.0, 9.2) | 5.2 (4.4, 6.1) | <0.01 | 1-0: <0.01 2-0: 0.14 2-1: 0.03 |
| PCSI cognitive | 12.6 (11.6, 13.7) | 8.9 (7.1, 10.8) | 15.5 (13.2, 17.8) | 13.6 (12.1, 15.0) | <0.01 | 1-0: <0.01 2-0: <0.01 2-1: 0.46 |
| Returned to exercise?b No. (%) | <0.01 | 1-0: 0.09 2-0: <0.01 2-1: <0.01 |
||||
| Yes | 51 (15.4%) | 31 (33.7%) | 13 (20.3%) | 7 (4.0%) | ||
| No | 267 (80.4%) | 59 (64.1%) | 47 (73.4%) | 161 (91.5%) | ||
| N/A (did not exercise before injury) | 14 (4.2%) | 2 (2.2%) | 4 (6.2%) | 8 (4.5%) | ||
| Returned to sport?b No. (%) | <0.01b | 1-0: 0.21c 2-0: <0.01c 2-1: <0.01c |
||||
| Yes | 20 (6.0%) | 13 (14.3%) | 5 (7.8%) | 2 (1.1%) | ||
| No | 271 (81.6%) | 66 (71.7%) | 52 (81.2%) | 153 (86.9%) | ||
| N/A (did not play organized sports before injury) | 40 (12.1%) | 12 (13.0%) | 7 (10.9%) | 21 (11.9%) | ||
| Not Reported | 1 (0.3%) | 1 (1.1%) | 0 (0.0%) | 0 (0.0%) | ||
| Returned to school?b No. (%) | <0.01 | 1-0: 0.82 2-0: <0.01 2-1: <0.01 |
||||
| Yes | 209 (63.0%) | 73 (79.3%) | 50 (78.1%) | 86 (48.9%) | ||
| No | 82 (24.7%) | 13 (14.1%) | 7 (10.9%) | 62 (35.2%) | ||
| N/A (on vacation) | 41 (12.3%) | 6 (6.5%) | 7 (10.9%) | 28 (15.9%) |
Two participants did not complete PCSI-SR13, thus PCSI values are based on N=330 and N=174, respectively.
Only Yes and No categories included in analysis
Exact tests used
In the overall sample of adolescents, a relatively small proportion reported having returned to exercise (15.4%) or organized sports (6.0%) prior to their initial clinical visit. There was an overall pattern that those who reported “Driving with No Changes,” had a higher proportion of adolescents having returned to exercise or sport compared to those who reported “Driving with Changes” or “Has Not Driven Since Injury” (exercise: 33.7% vs 20.3% vs. 4.0%; sport 14.3% vs 7.8% vs. 1.1%) (Table 3).
In the overall sample, 63% (n=209) self-reported having returned to school prior to their initial clinical visit. There was a significant difference in the proportion of adolescents who had returned to school among the three groups of post-injury driving behavior. The group “Has Not Driven Since Injury” had the lowest proportion of adolescents who had returned to school (48.9%), while the group “Driving with No Changes” had the highest proportion (79.3%) (Table 3). At the end of the initial clinical visit, for adolescents who self-reported that they were still in the school year (n=291), the clinical care team advised only 8.9% receiving advice to return without accommodations and most to return to school with accommodations (86.3%) (Table 4). The “Driving with No Changes” group had a significantly higher proportion of adolescents returning to full days of school without accommodations (23.3%) as compared to the “Driving with Changes” (1.8%) (p<0.01) and “Has Not Driven Since Injury” (3.4%) groups (p<0.01).
Table 4.
Recommendations for Return to School After Clinical Assessment
| Samplea (N = 291) | Driving with No Changes (N = 86) (0) | Driving with Changes (N = 57) (1) | Has Not Driven Since Injury (N = 148) (2) | p-value | Post-hoc | |
|---|---|---|---|---|---|---|
| Clinical Recommendation for Return to School,b No. (%) | <0.01b | 1-0: <0.01c 2-0: <0.01c 2-1: 0.57c |
||||
| Attend full days of school | 26 (8.9%) | 20 (23.3%) | 1 (1.8%) | 5 (3.4%) | ||
| Attend school with accommodations or engage in cognitive rest | 251 (86.3%) | 64 (74.4%) | 54 (94.7%) | 133 (89.8%) | ||
| Other | 9 (3.1%) | 1 (1.2%) | 1 (1.8%) | 7 (4.7%) | ||
| Not reported | 5 (1.7%) | 1 (1.2%) | 1 (1.8%) | 3 (2.1%) |
Of those who responded Yes (n=209) or No (n=82) on self-report Returned to School (NA n=41 not included here in this table)
Not reported not included in analysis
Exact tests used
Discussion
This study represents a unique examination of the self-reported driving practices of a large sample of adolescents during the recovery period after a concussion. Among adolescents aged 16-19 years diagnosed with a concussion in a specialty care setting, we found that close to half reported having returned to drive when asked approximately two weeks after injury, whereas few had returned to exercise and organized sports. Of those who returned to drive, over half reported not making any changes to driving behavior since the injury. Those that did make changes predominantly limited their number of driving trips. Adolescent drivers who had returned to drive, but had changed their driving behavior, endorsed a higher symptom burden than those who did not change their behavior. Among those driving without changes, three quarters were recommended cognitive rest or return to school with accommodations after the clinician’s assessment.
The percentage of adolescents who had refrained from driving post-injury in our study was similar to post-injury practices reported by adults in Preece et al.[24] (48%) and Schmidt et al.[18] (44%). However, Preece et al.[24] examined self-report of intentions to return to driving by adults ages 18-65 years seen in an emergency department within 24 hours of injury, and Schmidt et al. examined self-report data on previous concussions and return to drive in collegiate student athletes. The variation in return to drive post-injury in the Schmidt et al. study included time-based responses (e.g., driving <24 hours since injury, driving within 5-7 days), symptom-specific decisions (e.g., until symptoms were gone) and health care provider-imposed restrictions (e.g., until cleared by a health care provider). Although our analysis did not capture the time point at which the adolescents returned to driving (e.g., within 24 hours of injury), our results provide important information about what adolescents and families do in the absence of standardized clinical guidelines in comparison to the other age groups previously studied. The data suggest that a similar proportion of adolescents return to driving after concussion, even though they are novices compared to more experienced adults or college students studied in previous literature. The adolescents who had not returned to drive yet had on average two fewer days since injury than those returning to driving with changes, indicating that time since injury is likely an important factor in the process of returning to drive. The two-week mark may be an important time point the relationship, such that even an additional day or two post-injury may influence an adolescent’s return to the road. This deserves further exploration in a prospective study design.
Among the adolescents in our sample, a little less than half had returned to drive, but only 15% had returned to exercise and 6% to sports. In addition, after the initial clinical assessment by the provider, almost 9 out of 10 adolescents were advised to return to school with accommodations or to engage in cognitive rest, with less than 1 in 10 receiving advice to return to school without accommodations. One quarter of the adolescents who had reported return to driving without any changes were advised to return to full school days with no accommodations, representing the highest proportion of the driving subgroups. Returning to school too quickly or without accommodations can result in negative sequelae, such as cognitive, behavioral, academic, and/or emotional problems.[25] Although we do not know about the consequence of return to drive too quickly or without adjustments, the value of evidence for graded return-to-learn with cognitive demands should be considered. In the classroom and school setting, this gradation is designed mainly to mitigate symptom exacerbation. For driving, however, the exacerbation of symptoms does not represent the only risk. Premature return to drive in the context of cognitive, oculomotor and neurophysiological deficits of concussion may also place the adolescent, and the road-users around them, at risk for motor vehicle crashes. This may be particularly relevant for adolescents, as they are already the highest risk for motor vehicle crashes.[26]
Our analysis points to potential self-regulation occurring among the adolescents regarding activities and symptoms. A lower proportion of those that modified their driving (either they were not driving or they were driving with changes) were back to exercise and sport. Among those who had returned to driving but were making changes, concussion symptom burden was higher than those who were not making changes, indicating the interrelationship of symptoms and behavior modification. The modifications varied from not yet returning to driving to limiting the number of trips. Due to the cross-sectional nature of the analysis, we are unable to determine the direction of causality; for example, whether elevated symptom burden was the cause for changes in driving behavior, or whether engaging in driving behavior provoked symptoms. Since symptoms and clinical assessment have been reported as key factors in recommendations to adolescents and families in the absence of evidence-based and standard guidelines,[27] it is key to understand the directionality of the relationship between driving changes and symptoms.[7,9] Future studies should also assess the basis for the decision-making of adolescents and parents in their choices to refrain from driving or change driving behavior post-injury.
In a driving simulator study comparing asymptomatic young adults with a history of concussion with matched controls, Schmidt et al.[17] found that those with a history of concussion had poorer vehicle control. Other studies have shown that individuals diagnosed with a concussion are slower to identify driving hazards.[28,29] Although we did not evaluate driving performance on the road, the changes these adolescents were making related more to decreasing their driving exposures—limiting number and length of trips—rather than efforts to modify specific vehicle control skills, such as increasing space between other cars or limiting lane changes. Further quantification of these changes in exposure and skills using methods such as driving simulation, surveys or on-road monitoring of driving behavior is needed to better determine how to develop standardized guidelines to support adolescent driving safety after a concussion.
This study is not without limitations. The patients in our sample were all seen in a specialty care concussion program approximately two weeks after their injury, thus our results may not be generalizable to other sub-groups of concussion patients, such as those seen in the emergency department or urgent care, or those at different time points in their recovery. Data on whether the adolescent was a permit (learner) driver, licensed driver, and length of licensure was not available in the EHR. Given their importance in understanding adolescent crash risk, [30–32] future studies should collect this information and examine relationships with resumption of driving after concussion. This cross-sectional analysis of data from the Minds Matter Concussion Registry should help to inform future prospective studies that objectively measure driving behaviors and link them with functional neurocognitive deficits associated with concussion are needed.
Conclusion
Among adolescents aged 16-19 diagnosed with a concussion in a specialty care setting, nearly half reported return to drive since the injury, whereas less than 20% had returned to exercise and less than 10% had returned to organized sports at approximately two weeks post-injury. However, in the absence of structured clinical guidelines for gradually increasing driving activity, only 41% of adolescents who had returned to drive had modified their behavior. Four out of five adolescents were advised to return to school with accommodations to begin a gradual increase in school-related cognitive activities without severely exacerbating symptoms, suggesting that a similar gradual increase in the cognitive activities of driving may also be justified. While our results better characterize driving behavior changes made by adolescents after concussion, further research is needed to identify the objective metrics that can be measured clinically that define the stages and timing of an evidence-based return to drive protocol.
Implications and Contributions:
Many adolescents continued to drive after concussion despite having not yet returned to full exercise, sport or full cognitive activity/school. Nine out of ten were advised to return to school with accommodations to begin a gradual increase in cognitive activity, suggesting a gradual increase in driving may be justified.
Acknowledgements:
Research reported in this publication was supported by the National Institute of Nursing Research of the National Institutes of Health under Award Number R01NR018425. This material is also based upon work supported by the National Science Foundation under Grant Number EEC-1460927. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health or the National Science Foundation. The authors thank members of the Children’s Hospital of Philadelphia Sports Medicine and Performance Center Team including Naomi Brown, MD, FAAP, CAQSM, Matthew Grady, MD, FAAP, CAQSM, Christopher Renjilian, MD and Brian Vernau, MD, FAAP, CAQSM.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- [1].Bryan MA, Rowhani-Rahbar A, Comstock RD, et al. Sports-and recreation-related concussions in US youth. Pediatrics 2016;138:1–8. [DOI] [PubMed] [Google Scholar]
- [2].Sroufe NS, Fuller DS, West BT, et al. Postconcussive Symptoms and Neurocognitive Function After Mild Traumatic Brain Injury in Children. Pediatrics 2010;125:e1331–9. [DOI] [PubMed] [Google Scholar]
- [3].Vernau BT, Grady MF, Goodman A, et al. Oculomotor and neurocognitive assessment of youth ice hockey players: Baseline associations and observations after concussion. Dev Neuropsychol 2015;40:7–11. [DOI] [PubMed] [Google Scholar]
- [4].Thiagarajan P, Ciuffreda KJ, Ludlam DP. Vergence dysfunction in mild traumatic brain injury (mTBI): A review. Ophthalmic Physiol Opt 2011;31:456–68. [DOI] [PubMed] [Google Scholar]
- [5].Thornhill S Disability in young people and adults one year after head injury: prospective cohort study. BMJ 2000;320:1631–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [6].Marshall S, Bayley M, Mccullagh S, et al. Clinical practice guidelines for mild traumatic brain injury and persistent symptoms. vol. 58. 2012. [PMC free article] [PubMed] [Google Scholar]
- [7].Klein TA, Graves JM, Graham JY. Driving After adolescent concussion: Advice from nurse practitioners in the absence of standardized recommendations. J Pediatr Heal Care 2017;31:441–5. [DOI] [PubMed] [Google Scholar]
- [8].Lucas JA, Moore JB, Davis S, et al. Provider attitudes and management regarding returning to drive after concussion. Br J Sports Med 2019;53:495. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [9].MacDonald J, Patel N, Young J, et al. Returning adolescents to driving after sports-related concussions: What influences physician decision-making. J Pediatr 2018;194:177–81. [DOI] [PubMed] [Google Scholar]
- [10].Baker A, Unsworth CA, Lannin NA. Determining fitness to drive: A systematic review of the methods and assessments used after mild traumatic brain injury. Br J Occup Ther 2015;78:73–84. [Google Scholar]
- [11].Centers for Disease Control and Prevention. Teen drivers: Get the facts. Available at: http://www.cdc.gov/Motorvehiclesafety/teen_drivers/teendrivers_factsheet.html. Accessed June 18, 2020.
- [12].Centers for Disease Control and Prevention. Web-based Injury Statistics Query and Reporting System (WISQARS): Leading causes of nonfatal injury reports. Available at: http://www.cdc.gov/injury/wisqars/. Accessed June 18, 2020.
- [13].Insurance Institute for Highway Safety. Teenagers: Overview. Available at: http://www.iihs.org/iihs/topics/t/teenagers/topicoverview. Accessed June 18, 2020.
- [14].Preece MHW, Horswill MS, Ownsworth T. Do self-reported concussions have cumulative or enduring effects on drivers’ anticipation of traffic hazards? Brain Inj 2016;30:1096–102. [DOI] [PubMed] [Google Scholar]
- [15].Baker A, Unsworth CA, Lannin NA. Fitness-to-drive after mild traumatic brain injury: Mapping the time trajectory of recovery in the acute stages post injury. Accid Anal Prev 2015;79:50–5. [DOI] [PubMed] [Google Scholar]
- [16].Bottari C, Lamothe M-P, Gosselin N, et al. Driving Difficulties and Adaptive Strategies: The Perception of Individuals Having Sustained a Mild Traumatic Brain Injury. Rehabil Res Pract 2012;2012:1–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [17].Schmidt JD, Hoffman NL, Ranchet M, et al. Driving after Concussion: Is It Safe To Drive after Symptoms Resolve? J Neurotrauma 2017;34:1571–8. [DOI] [PubMed] [Google Scholar]
- [18].Schmidt JD, Lynall RC, Lempke LB, et al. Post-concussion driving behaviors and opinions: A survey of collegiate student-athletes. J Neurotrauma 2018;35:2418–24. [DOI] [PubMed] [Google Scholar]
- [19].Foss RD, Martell CA, Goodwin AH, et al. Measuring changes in teenage driver crash characteristics during the early months of driving. AAA Foundation for Traffic Safety; Washington, DC: 2011. [Google Scholar]
- [20].McCrory P, Meeuwisse W, Dvořák J, et al. Consensus statement on concussion in sport-the 5th international conference on concussion in sport held in Berlin, October 2016. Br J Sports Med 2017;51:838–47. [DOI] [PubMed] [Google Scholar]
- [21].Sady MD, Vaughan CG, Gioia GA. Psychometric characteristics of the postconcussion symptom inventory in children and adolescents. Arch Cinical Neuropsychol 2014;29:348–63. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [22].Frommer LJ, Gurka KK, Cross KM, et al. Sex Differences in Concussion Symptoms of High School Athletes. J Athl Train 2011;46:76–84. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [23].Bergdahl J Sex differences in attitudes toward driving: A survey. Soc Sci J 2005;42:595–601. [Google Scholar]
- [24].Preece MHW, Geffen GM, Horswill MS. Return-to-driving expectations following mild traumatic brain injury. Brain Inj 2013;27:83–91. [DOI] [PubMed] [Google Scholar]
- [25].O’Neill J, Cox M, Clay O, et al. A Review of the Literature on Pediatric Concussions and Return-To-Learn (RTL): Implications for RTL Policy, Research, and Practice. Rehabil Psychol 2017;62:300–23. [DOI] [PubMed] [Google Scholar]
- [26].Centers for Disease Control and Prevention. Ten Leading Causes of Death and Injury. Available at: http://www.cdc.gov/injury/wisqars/leadingcauses.html. AccessedJuly 15, 2020.
- [27].Schmidt JD, Lempke LB, Devos H, et al. Post-concussion driving management among athletic trainers. Brain Inj 2019;33:1652–9. [DOI] [PubMed] [Google Scholar]
- [28].Hoffman NL, Devos H, Schmidt JD. Driving performance deficits despite concussion symptom resolution: A case study. Int J Athl Ther Train 2018;23:21–6. [Google Scholar]
- [29].Preece MHW, Horswill MS, Geffen GM. Driving after concussion: The acute effect of mild traumatic brain injury on drivers’ hazard perception. Neuropsychology 2010;24:493–503. [DOI] [PubMed] [Google Scholar]
- [30].Mayhew DR, Simpson HM, Pak A. Changes in collision rates among novice drivers during the first months of driving. Accid Anal Prev 2003;35:683–91. [DOI] [PubMed] [Google Scholar]
- [31].Williams AF. Graduated driver licensing (GDL) in the United States in 2016: A literature review and commentary. J Safety Res 2017;63:29–41. [DOI] [PubMed] [Google Scholar]
- [32].Curry AE, Pfeiffer MR, Durbin DR, et al. Young driver crash rates by licensing age, driving experience, and license phase. Accid Anal Prev 2015;80:243–50. [DOI] [PubMed] [Google Scholar]