Advancing Understanding of Transportation Options

For the values clarity subscale, three decision conflict scale (DCS) items (Likert scale responses from 0 ['strongly agree'] to 4 ['strongly disagree']) are summed, divided by 3 and multiplied by 25; scores range from 0 (extremely clear) to 100 (extremely unclear about personal values); lower scores represent higher values clarity, which is a better outcome. Lower values clarity subscale scores are associated with positive outcomes. The measure will be assessed in drivers immediately following administration of control condition or intervention.

The following true/false questions will be used to examine participants' safe driving knowledge: 1) Older drivers pose a bigger risk to other drivers and pedestrians than younger drivers do; 2) Drivers aged 70 years and older are more likely to be hurt or killed if they are in a crash; 3) There is a milestone age when everyone should stop driving; 4) With age, a person needs more light to see well; 5) Getting lost on familiar roads is a sign that it might be time to stop driving. A higher number of correctly answered questions indicate better outcomes. Knowledge will be defined as the percentage of correct scores of the 5 questions (ranging from 0% [worst outcome] to 100% [best outcome]). The measure will be assessed in drivers immediately following administration of control condition or intervention.

The Decision Self-Efficacy Scale will be used to assess decision self efficacy. Scores range from 0 (extremely low self-efficacy; worst outcome) to 100 (extremely high self-efficacy; best outcome). Higher scores indicate better outcomes. The measure will be assessed in drivers immediately following administration of control condition or intervention

Depression will be measured using the PROMIS Short Form 4-item scale (4a Adult v1.0), with higher PROMIS scores indicating higher depression. PROMIS scores are presented as standardized T-scores (mean=50, standard deviation=10). Lower PROMIS depression scores indicate better outcomes (lower depression).

Change in depression is calculated as the PROMIS depression score at each time point (6, 12, 18, or 24 months) minus the score at baseline (pre-randomization). Since PROMIS scores at each time point can range from 41.0 to 79.4, change in depression PROMIS scores can range from -38.4 to +38.4. Change-scores > 0 represent increases in depression (poor outcome), and changes <= 0 represent maintained or decreased depression (better outcome).

Change in decision regret will be measured by the Ottawa Decision Regret Scale. This validated measure correlates with decision satisfaction and conflict, and overall quality of life. Scored from 0-100, high scores represent higher regret. Lower scores indicate better outcomes.

Decision regret was measured at 6, 12, 18, and 24 months, so change in decision regret can be calculated from 12 months (vs. 6 months), 18 months (vs. 6 months), and 24 months (vs 6 months). Since decision regret scores range from 0-100, change in decision regret scores range from -100 to + 100, with change-scores <= 0 representing maintenance or decreases in decision regret (positive outcome) and change-scores > 0 representing increased decision regret over time (negative outcome).

Life space is a global measure of mobility and community engagement. The Life-Space Assessment instrument (UAB Study of Aging) is a validated tool assessing recent mobility and function. Composite scores range from 0 (bedbound) to 120 (travel out of town every day without assistance); scores of ≤60 are correlated with lower levels of social participation and higher mortality. Higher scores indicate better outcomes.

Change in life space can be calculated at 6 months (vs baseline), 12 months (vs baseline), 18 months (vs baseline), and 24 months (vs baseline). Since life space scores range from 0-120, change-scores in life space can range from -120 to +120. Maintained or increased life space (change-scores >= 0) are positive outcomes, and decreased life space (change-scores < 0) are negative outcomes.

Self-reported driving frequency will be measured by number of days per week participants drive, with consideration of higher or lower frequency as a positive or negative outcome in the context of participant's intent to drive with lower or higher frequency.

In the final study survey, this question was reworded, so participants were asked "during the past 6 months, have you reduced the number of days per week you normally drive?" with response options of "yes", "no", "I don't know". Percentage of those who answered "yes" at each time point are reported.

Driving avoidance in certain situations will be measured by asking participants about their avoidance of riskier driving situations, e.g., nighttime driving, with consideration of avoidance in certain situations as a positive or negative outcome in the context of participant's intent to avoid these situations.

Avoidance of night driving was worded as: "Using a scale from 1 to 7 where 1 is Not At All Comfortable and 7 is Completely Comfortable, how comfortable do you feel in the following situation?", and the situation presented was "Driving at night?". Higher scores indicate greater comfort with driving at night, and lower scores represent less comfort. Therefore, change-scores can range from -6 to + 6, with change-scores >= 0 indicate maintained or increased comfort with driving at night, and change-scores < 0 indicate reduced comfort in driving at night.

Driving cessation will be measured by asking participants whether they are currently driving; all participants were currently driving at study baseline. Consideration of driving cessation as a positive or negative outcome will be determined in the context of participant's intent to engage in driving cessation. This question was worded as "Do you currently drive?" with response options of "yes" and "no". The numbers reported are those who said "no", which represents driving cessation.

The wording of this question was added to the study after an early DSMB meeting, so less data are available for it at earlier time points.

Occurrence of self-reported crashes will be measured by ≥1 crash versus no crashes, where ≥1 crash is a negative outcome.