BACKGROUND:

Despite widespread application in aviation and other fields, there has been limited use of computerized simulation in driver education. We prospectively studied a group of novice drivers subjected to comprehensive virtual driving simulation modules to identify the subsequent effects on their driving records. We hypothesized that participation in a simulation program would result in fewer offences and crashes. METHODS:

Forty high school students who recently obtained their driver's license were randomized into driving simulator (DS) or control groups. The DS group went through 12 modules of driver education. Upon completion, driving records for all the individuals were collected at 6 months, 12 months, and 18 months, and comparisons were made. Statistical analysis was performed using χ², Fisher's exact tests, t tests, and Mann Whitney U-test where appropriate. RESULTS:

Of the 20 subjects, 16 in the DS group completed all modules and were compared with 19 individuals in the control group. Sixty-nine percent in the DS group were male versus 89% in the control group. Mean age was similar in both groups. The average time to the first offense after completion in the DS group was 117 days versus 105 days in control group (p = 0.8). At 18 months, 18.8% in the DS group were involved in a driving incident compared with 47.4% in the control group (p = 0.1516). At 18 months, there were 4 incidents (0.25 incidents per person) in the DS group versus 17 incidents (0.89 incidents per person) in the control group. At 18 months, 6.2% in the DS were involved in accidents compared with 21.1% in the control group (p= 0.35). Speeding infractions occurred at 18 months in 12.5% in the DS group versus 26.3% in the control group (p = 0.4150). CONCLUSION:

In this prospective pilot evaluation of computerized driving simulation, adolescents subjected to structured simulator training showed trends toward committing fewer offences and accidents. Larger studies examining the practical potential of driving simulation in novice drivers are needed. LEVEL OF EVIDENCE:

Prognostic study, level III.