Background Simple and valid measures are important for studying distracted walking, an important and emerging risk factor for pedestrian injury. Based on common epidemiological indicators for distracted walking, we developed and validated an indicator to measuring distracted walking, called the ‘distraction index’.

Methods A large, video-based, observational study was conducted at 20 intersections in Changsha, China. To develop a ‘distraction index’ that would quantify safety risks based on distracted walking, we calculated relationships between 7 observed distraction indicators and 3 safety outcomes. This allowed us to assess the discriminant validity of the 7 indicators and create a ‘distraction index’ that best predicts street-crossing safety based on distracted behavior.

Results In total, 8,729 distracted pedestrian crossings were identified. Discriminant validity varied greatly across the distraction indicators. Compared to each individual indicators, the distraction index demonstrated the strongest capacity to discriminate all three safety measures of street-crossing. Using the three levels in the ‘distraction index’, distracted pedestrians had much higher risks of experiencing near-crash events compared to undistracted pedestrians, with odds ratios of 1.3 (95% CI: 1.1, 1.4) for low index scores, 1.6 (95% CI: 1.4, 1.9) for medium, and 1.9 (95% CI: 1.6, 2.3) for high.

Conclusion The distraction index predicted pedestrian crossing safety more accurately than any of the seven individual measures of distracted walking. We recommend use of this index in future research.

Learning Outcomes No widely recognized epidemiological measure exists to observe distracted walking. The newly-created ‘distraction index’ shows excellent discriminant validity compared to individual indicators of distraction.