A population-based case-control study is proposed. The study base includes all drivers of heavy vehicles currently in operation in 3 Australian states namely, New South Wales (NSW), Western Australia (WA) and Queensland (QLD) over a 3-year study period. Drivers will be included if they are undertaking long distance trips (defined as trips in excess of 200 km from their work base), and driving a vehicle ≥12 tonnes. Whilst buses fit this weight criterion, they are excluded from participation in the study due to the significant difference in their scheduling and travel patterns compared to those drivers under study.
Drivers of heavy vehicles that are involved in a crash in one of the above-mentioned states during the 3-year study period will be eligible for recruitment into the study. All crashes must be attended by police and therefore, captured in the respective state police databases. Systematic search strategies have been established that will be run weekly by police across these crash databases. The weekly reports will be sent (via security encrypted files) to the research team, who then send study information and letters of invitation to participate in the study to each heavy vehicle driver that meets the study inclusion criteria. Drivers will be excluded where they have been seriously or fatally injured, or where the crash has resulted in the fatal injury of another road user. Serious injury to the driver is defined as that which 'requires hospitalisation for more than a 1 week period', or the heavy vehicle driver was transported 'unconscious' from the scene of the crash. Telephone contact with those drivers receiving the study information will then be undertaken two weeks from the date the letter was sent. Those drivers not wishing to participate in the study are able to decline via a return mail or alternatively, at the point of telephone contact by the research team. For participating drivers, once informed consent is obtained, a 40 minute telephone interview will be conducted.
Heavy vehicle drivers (as per the definition above) who have not been involved in a police-reported crash in the 12 months leading up to the time of interview will be eligible for recruitment into the study. All refuelling station/truck stop sites across each state have been identified, and heavy vehicle travel patterns across the road network in each state have been obtained from the respective state roads jurisdictions. Using these details, a range of refuelling station/truck stops have been randomly selected to capture a representation of heavy vehicles travelling across the road network in each state. The number of locations selected per state ranges from 6 to 15 sites.
The interviews will be conducted in blocks of time from 4-8 hour time-periods and across the days of the week to ensure weekly travel patterns of heavy vehicle drivers are captured. Once the heavy vehicle refueling station/truck stop and interview time/day schedule is selected, the interviewers will approach heavy vehicle drivers entering the refueling station/truck stop to participate in the study. Interviewers will continue recruiting at each site until the available or recommended time period has elapsed. Demographic information such as age, height, weight and loads carried will be requested of non-participants.
Both case and control interviews are conducted by a consistent group of staff all trained using a standardized protocol. As well, since the diagnosis of sleep apnoea among the case subjects could be sensitive and may have ramifications for their ongoing employment, we will be emphasising strongly the confidential nature of the study. Given the importance of maintaining a sufficient response fraction, rigorous attention will be paid to recruitment strategies and the development of rapport with potential participants. However, in anticipation of those drivers who decline participation in the study, and to minimise the potential for bias, we will collect demographic data on these drivers. Further, ensuring methodological rigour, we are using validated diagnostic measures for potential risk factors where possible.
Driver Interview and Sleep Apnoea testing
A questionnaire was developed by consensus of a panel comprising experts in transportation, the heavy vehicle industry, fatigue and sleep disorders, legislation, injury prevention, and road safety. The questionnaire collects information about the driver's characteristics including demographics, anthropometry, health status, drug and alcohol use, sleep quantity and quality and fatigue. For self reported measures of sleepiness, the questionnaire includes standardised, validated questions such as the Epworth Sleepiness Scale , and Multivariable Apnoea Prediction Index . Summary information from all drivers will include their driving and crash history, income details, distances travelled and hours worked. Vehicle information obtained will include truck weight and configuration, safety features and typical loads carried. Detailed information will be further collected about work schedules, travel patterns and sleep (quantity and quality) for the three days preceding the crash (for cases) or the three days preceding the face to face interview (for controls).
In order to test for obstructive sleep apnoea, all drivers will be sent a diagnostic device (Flow Wizard™) to wear overnight during their sleep. This device is a validated nasal pressure transducer that non-invasively measures an individual's airflow during sleep and will detect apnoeas, hypopnoeas and snoring per hour of recording [20, 21]. It can be worn with minimum instruction and will be mailed back to the research team for analysis. The device has shown reliable accuracy in detecting obstructive sleep apnoea (sensitivity of 0.92), and with sufficient data can classify results as 'no/minimal', 'moderate' or 'severe' disease .
A total sample of 1034 heavy vehicle drivers will be recruited comprising 517 cases and 517 controls. A number of studies have shown that the risk of crashing for heavy vehicle drivers doubles with increased hours of driving . Drivers in a New Zealand study were 2.6 times more likely to crash if they had driven 8 or more hours since their last compulsory 10 hour break, compared with those who had driven less than 8 hours . Similarly, research in the United States reported a 2.2 fold increase in the likelihood of crash for each additional 100 km travelled . With regard to sleep/fatigue related covariates, Santos et al  reported an association between a diagnosis of sleep apnoea and the risk of crash among car drivers, and hours of sleep over the previous day have been associated with an increased risk of crash in car drivers (< 5 hrs, OR 2.7) as well as those driving between 2-5 am (0.4%, OR 5.6) . In the absence of research identifying the association between specific schedules/payment and crash, the power for the study was estimated on associations between heavy vehicle crash and sleep related conditions. Therefore taking a conservative prevalence estimate of 5%, with an expected odds ratio of 2.0, the minimum sample size necessary to detect a true odds ratio with 80% power is 517 cases and 517 controls.
Descriptive statistics will profile the study population on all demographic data and survey responses, and then comparisons will be made between the cases and controls. A series of univariate analyses will be conducted using chi square tests for categorical variables and t tests for continuous variables. Factors that appear to be associated with crash risk will be included in a multivariate model, determining the odds of crash. The odds ratio will be estimated related to the key aims of the study namely, factors about payment, hours worked, shift scheduling and a diagnosis of sleep apnoea. Unconditional logistic regression analysis will be undertaken, adjusting for confounding factors. For all tests, probability values of less than .05 will be considered statistically significant. Odds ratios and 95% confidence intervals will be derived using the standard case-control study methods described by Breslow and Day .
The control group of 517 drivers will then be analysed separately, (as a prevalence survey) to describe the prevalence of key risk factors among heavy vehicle drivers. Statistical calculations will be performed on STATA software (version 10.0 Stata Corporation, College Station, TX, USA).
Informed Consent and Ethics
Informed consent will be obtained from all participants; control drivers will be read a consent statement after provision of a description of the study, and then will sign the consent form (with a second copy for their records). For case drivers, where interviews are conducted over the telephone, they will be informed about the study, and a supervisor of the interviewer will listen to the consent process on a second telephone line. The supervisor will confirm the consent process in writing.
Ethics approval for the study was been obtained in January 2008 from The University of Sydney Human Research Ethics Committee; prior to approval by Curtin University Human Research Ethics Committee in WA and the Ethical Standing division of Queensland Police in Qld. All applications for amendment with refined versions of Participant Information Statements, Interviews, Consent forms etc. have been approved.