Document resource
Introduction Untreated sleep apnoea is known as a risk factor for road traffic collisions. Nonetheless, while collisions are more frequent than for people without apnoea, they are still rare and there are no evidence-based markers to assess this risk. Motion detection is used by insurers to assess risky driving. We sought to establish if similar phone generated measures of driving were different in people with sleep apnoea and those without.Methods Participants were recruited at referral to our sleep service. After informed consent they uploaded a phone app (Insights app, Sentiance Belgium) and recorded driving data pre and post diagnosis and for a sub-group post treatment with mandibular devices or CPAP. Data were analysed for events of acceleration, braking and turning using linear mixed effects models against participant data including age, sex, Epworth score (ESS), desaturation index (ODI) and length of and time into journey.Results Ninety-one participants, mean age 51 (22 to 77) years, 35% female, mean ESS 12.5, mean ODI 15/hr were recruited and 7,834 trips, mean duration 15.6 minutes with 508,544 ‘events’ were analysed. ESS and ODI as independent factors were not related to driving behaviour while younger age showed riskier behaviour. However grouping participants using both ESS and ODI showed that compared to non-sleepy people (ESS< 11) with no sleep apnoea [no SA = ODI <5), the groups: sleepy & no SA, sleepy & mild SA, sleepy & mod to severe SA, were all identifiably different especially in measures of turning magnitude while non-sleepy people with moderate to severe SA were not.Discussion Telematic driving data differentiate people with self-reported sleepiness and sleep apnoea from non-sleepy individuals without sleep apnoea. Further analysis may yield practical measures to improve risk assessments for crashes and give objective support for decisions around advice on driving cessation or continuance.