The effect of non-infrastructural visibility aids on crashes and injuries among commercial transport pedal bicyclists in Eldoret

Abstract

Background: Effectiveness of non-infrastructural visibility aids on bicycle crash injury prevention was not well established, yet they were widely used. Objective: To determine influence(s) of non-infrastructural visibility aids on specified riders’ crashes and injuries within defined low visibility periods of the 24 hour day. Setting: Eldoret Municipality, Kenya. Study Design: Cross-sectional survey. Methods: A pre tested semi structured, interviewer administered questionnaire was used on 364 riders, to determine their self reported characteristics and crash injury details for the year 2008. Non infrastructural visibility aids on riders and their bicycles were determined by an observation checklist. Data entry, cleaning and analysis employed the Computer program, Epi_Info version 3.5.1. Frequencies of variables were determined. Two by two tables and multivariate logistic regression, determined significant relations between specific non infrastructural visibility aids and riders’ crashes or injuries, specified by both defined crash periods (night, non-daytime or whole day) and received management (self care, outpatient care or in-patient care). Outputs were: Odds ratios, 95% confidence intervals, coefficients and p-values. Presentations were in summary measures and tables. Results: Riders’ main characteristics: 100% (N=364) male, 70.9 % (N=258) age group 21-30 years, 80.8% (N=294) 4 years or less riding experience, 75.0% (N=273) rode at least once at night, 99.2% (N=361) spoke Swahili and 90.4% (N=329) owned the observed bicycle boda boda. Proportions of bicycle boda boda with specific visibility aids: 62.9% (N=229) rear reflectors and 27.2% (N=99) working headlight. Riders’ proportions with reflective strips and group uniform were 55.5% (N=202) and 69.2% (N=252) respectively. Riders’ injuries were 77.7% (N=369). Self involved riders’ injuries were the most common, 54.2% (N=200), followed by motor vehicle related ones, 27.9% (N=103). Among injuries managed by in-patient care, 60.0% (N=18) occurred within nondaytime hours. Bright garments were associated with 34% reduction of non-daytime crashes (OR=0.6577, 95%CI: 0.4539-0.9530). At multivariate level, working headlight was associated with 75% reduction (OR=0.2539, 95%CI: 0.0771-0.8358) of non-daytime injuries which were managed by outpatient care. Conclusion: Non-infrastructural visibility aids may be useful in prevention of injuries within periods of the 24 hour day with low average natural visibility.