Article Text


Assessing the impact of interventions in sleep-disordered breathing
S15 An office based advanced driving simulator to assess driving performance in Obstructive Sleep Apnoea Syndrome (OSAS): a pilot study
  1. D Ghosh1,
  2. S Jamson2,
  3. M W Elliott1
  1. 1St James’ University Hospital, Leeds, UK
  2. 2Institute for Transport Studies, University of Leeds, Leeds, UK


Introduction Advising patients with Obstructive Sleep Apnoea Syndrome (OSAS) about whether they are safe to drive is challenging. Driving simulator studies have shown that OSAS patients perform poorly (Hack et al, 2001) but most simulators are simple, not realistic and in particular lack face validity, for example, multiple off road events during a short run. The Institute for Transport Studies, University of Leeds, host the UK's most sophisticated driving simulator but had also developed a PC based simulator (MiniSim) which incorporates the same realistic graphics and is much closer to ‘proper’ driving than most existing systems. It thus has the potential to be used in everyday clinical practice. We have investigated whether two parameters, proportion of high frequency steering activity (HFS) and standard deviation of lane position (SDLP), previously shown to be impaired in drivers suffering from fatigue, might predict drivers’ behaviour in safety-critical scenarios.

Methods After a practice run, 63 patients (age 53±10, ESS 11±3, ODI 39±19) completed 50 min motorway driving on the MiniSim. Two situations were programmed that required evasive action to avoid a crash. A ‘fail’ was determined by an outright crash or veering completely out of lane. We compared HFS and SDLP in subjects with ‘pass’ or ‘fail’ and with Oxygen Desaturation Index (ODI) and Epworth Sleepiness Score (ESS).

Results (Abstract S15 Table 1) ‘Fail’ was more likely with worse sleep disordered breathing, but was not affected by subjective sleepiness. Subjects who ‘failed’ had significantly worse HFS (0.27 vs 0.34, p=0.03) & SDLP (0.58 vs 0.41, p=0.002). Both HFS & SDLP worsened with time. There was also a strong correlation between HFS & SDLP (r=0.51, p<0.0001).

Abstract S15 Table 1

Comparison between patients passing & failing the simulator run (Mean HFS=mean for the proportion of high frequency steering activity in epochs 3,6,7 of the simulator run, Mean SDLP= mean for the standard deviation of lane position in epochs 3,6,7 of the simulator run)

Conclusion These data show that HFS and SDLP have promise as objective markers of poor driving in OSAS patients. They relate to measures of disease severity and to an event which has face validity as an indicator of poor driving in the real world. The patient will not be aware that they are being measured and therefore they have potential for repeated use.

Statistics from

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.