Drinking and driving in an urban setting in Kenya.

A roadside alcohol prevalence survey of drivers randomly selected from the general traffic was conducted in Eldoret, Kenya. Blood alcohol concentration (BAC) data obtained by a breath test in 90% of the sample (n = 479) was analysed by demographic and travel characteristics. 19.9% had a positive breath test (BAC > or = 5 mg%), 8.4% had BACs greater than 50 mg%, and 4% exceeded 80 mg%. A greater proportion of males (20%) had been drinking compared to females (12.5%): all drivers with high BACs (> or = 50 mg%) were males. The likelihood of having consumed alcohol was greater in motorists aged 25 years and above (20.4%) than in younger drivers aged 16-24 years (15.4%), their mean BACs were also more elevated (57 mg% versus 31 mg%). In comparison to operators of public service vehicles (PSV), people driving personal cars were more than twice as likely to have been drinking: with 21.9% being BAC positive against 10.8% (OR = 2.3; 95% CI, 1.0 to 6.3, p = 0.05). Educated individuals with skilled careers tended to indulge in drink-driving to a greater extent than professional drivers (operators of public transport, taxi and heavy goods vehicles), with BAC prevalence rates of 23.7% and 15.5%, respectively. Other circumstances influencing the probability of drink-driving were number of vehicle occupants, distance to destination, road location, time of the night and whether it was a weekend or weekday. These findings are discussed in relation to the potential for promotion of relevant deterrent measures, including the establishment of an appropriate BAC legal limit for drivers in Kenya.