Is there a case for screening commercial drivers for sleep apnea?

Obstructive sleep apnea syndrome (OSAS) is a common condition associated with an increased risk of motor vehicle collisions (MVCs). Effective treatment is available, most commonly with continuous positive airway pressure (CPAP). Inexpensive screening tools are available that can effectively identify severe OSAS. However, other sleep-related disorders (e.g., sleep restriction and circadian rhythm disorders), which are of equal or greater importance than OSAS in the causation of MVCs, may not be identified using these screening tools. Moreover, the published data regarding the positive effects of treatment of OSAS on MVCs are limited to domestic rather than commercial drivers. Finally, financial and ethical aspects must be considered before a widespread roll-out of an OSAS screening program.

Approximately 20–30% of all MVCs are sleep related [1]. In the USA, the cost of large truck crashes was estimated at US$24 billion in 1999 [2] and the cost of sleep-related MVCs among commercial drivers in Australia has been estimated at AUS$300 million [3]. Hence, the identification and management of conditions leading to sleepiness in commercial drivers is clearly important.

Sleep apnea is a sleep-related disorder in which ventilation is not maintained adequately during sleep. The most common form of sleep apnea is obstructive sleep apnea (OSA), which results in episodic upper airway collapse, snoring, hypoxemia, hypercapnia and arousals from sleep associated with swings in blood pressure. OSAS is defined by at least five apneas or hypopneas (complete or partial reductions in ventilation, respectively, lasting at least 10 s) per hour, plus either symptoms (e.g., sleepiness) [4] or signs (e.g., systemic hypertension or heart failure). Several studies have identified a higher risk of MVC with untreated OSA in domestic [5–7] and commercial drivers [8–11]. A meta-analysis indicated that OSAS is associated with a two-to-threefold increased risk of MVC across a population of domestic and commercial drivers [12]. Reversal of OSAS with CPAP reduces the MVC risk to that of normal controls [5].

Sleep apnea is common. In 1992, it was estimated that 24% of male and 9% of female community dwellers had OSA (apnea/hypopnea index [AHI] > five episodes per hour [eph]) and 9% of males and 4% of females had OSA (AHI > 5 eph) plus symptoms [4], with similar numbers found elsewhere in the world [13]. A rising incidence of obesity and aging populations are likely to significantly increase these percentages. Some population groups (those with diabetes, obesity, atrial fibrillation, systemic hypertension on more than two drugs or heart failure) have a much higher prevalence of OSA, and are target groups for OSAS screening.

In the USA, Pack et al. reported that 28 and 5% of commercial drivers had OSA (AHI > 5 and > 30 eph, respectively) based on oximetry and polysomnography [9]. This confirmed previous observations in US commercial drivers [11]. In Australia, 60 and 11% of commercial drivers had OSA (AHI > 5 and > 30 eph, respectively) based upon polysomnography, whereas only 16% declared symptoms of OSA [8]. More importantly, 24% of Australian commercial drivers volunteered severe sleepiness (Epworth Sleepiness Scale [ESS] > 11 [14]), which was predicted statistically by OSA severity, in addition to other factors, such as the number of hours worked, afternoon or shift work plus stimulant or caffeine use [8]. There is a predominance of men, a high prevalence of obesity and large proportion of drivers aged over 40 years among commercial drivers in these countries that account for the high OSA prevalence [8]. The prevalence of OSA among commercial drivers appears to be lower in countries where these factors are not as prevalent [13].

Several mechanisms exist by which OSAS may contribute to MVCs. These include cognitive impairment secondary to intermittent hypoxemia, hypercapnia, fragmented sleep and reduced quality and quantity of sleep. Clinically, this is manifested by poor judgement, impaired memory, microsleeps, impaired hand–eye coordination, impaired arousal response to external stimulae (e.g., rumble strips), depression and suicide.

The increased risk of MVCs in OSAS subjects has been highlighted in nine studies, a meta-analysis of which suggests an overall two- to threefold elevated risk [12]. However, these studies have been performed in domestic (not commercial) drivers. In a study by Stoohs, the odds ratio of MVCs was not significantly greater in commercial drivers, however subjects numbers were small (n = 90) [10]. In a larger study by Howard et al. (n = 3268), commercial drivers with a very high level of OSAS had a slightly greater risk of single, but not multiple, MVCs [8]. No association between OSAS and MVCs was observed in a subgroup (n = 161) who underwent in-laboratory sleep studies.

Howard et al. also noted a greater risk factor for single and multiple MVCs with a high ESS questionnaire score (ESS > 18), as was estimated sleep duration of less than 7 h [8]. Thus, other sleep-related issues may be equally or more important than OSA in isolation.

This observation was recently confirmed in a prospective study of contributing factors for domestic drivers injured in MVCs sufficiently severely to require hospital admission [16]. Once alcohol, medical and psychiatric disease were excluded, shift work, and not OSA, was the main contributing factor to the accident.

In a study of commercial drivers who underwent 1-week actigraphy, detailed questionnaire, polysomnography or overnight oximetry, Pack et al. reported that 35% slept less than 6 h per night, 14% less than 5 h per night and 6% less than 4 h per night [9]. Moreover, 5% had severe OSA (AHI > 30 eph) and 28% mild OSA (AHI > 5 eph). Less than 5 h sleep per night was equivalent to severe OSA (AHI > 30 eph) when using the multiple sleep latency test (MSLT) and psychomotor vigilance reaction time (PVT) as tests of sleepiness.

Other factors that have been related to commercial MVCs include the road environment, shift length, vehicular factors and drug use [8,17,18]. The use of antihistamines and narcotic analgesics have been associated with an increased risk of single and multiple MVCs [8]. Other drugs, in particular stimulants, have been found to be related to risk of fatal MVCs in commercial drivers [19].

Thus, other important and prevalent causes of sleepiness include sleep deprivation (<8 h sleep), disturbed circadian rhythm (e.g., shift work), drugs (social and prescribed) and other conditions that might fragment sleep (arthritis, asthma or gastro–esophageal reflux disease [GORD]) or cause sleepiness (depression) need to be recognized.

Recognition of OSAS by history and physical examination alone is inaccurate, and this has led to attempts to use a combination of standardized questionnaires (e.g., ESS [14] and Berlin Questionnaire [20]), examination (e.g., neck circumference, Mallampati index and BMI), the multivariable apnea prediction (MAP) index [21] and a sleep study in high-risk groups. The MAP index incorporates a 13-point self-administered questionnaire of symptoms of OSAS and other sleep disorders, plus age, BMI and gender [21], and has been found to have reasonably good test–retest reliability (r = 0.92) and reliability of predicting OSA (AHI > 10 eph) with area under the ROC curve of 0.79 [21].

Recognition of OSAS can be made via a variety of monitoring techniques with the two ends of the spectrum being limited channel unattended studies at home (i.e., oximetry, airflow and respiratory effort) and more complex supervised (i.e., attended) polysomnographs (sleep state, ventilation and ECG) in a sleep laboratory. Of the two most important variables that influence OSA that require monitoring (namely sleep stage and body position), most attention to design and quality assurance has been paid to sleep stage. However, body position is the more important variable compared with sleep stage and, accordingly, should be accurately monitored [22].

Other tests of vigilance can complement the sleep study, such as PVT or mean wakefulness tests (MWTs) and MSLT, driving simulators and tests of circadian rhythm and sleep deprivation (actigraphy), plus urinary drug screening tests. These are important tests, required, at times, to assess the sleepy driver. However, data linking these measures to crash risk are limited and not convincing to those in the field, and are currently in the realm of research [7].

Treatment of OSAS with favorable outcome is available with a myriad of treatments: lifestyle (weight loss or restricted alcohol), surgical (nasal, palatal, maxillo–mandibular or weight loss), dental devices (maxillary expansion or mandibular advancement) and CPAP. The most commonly prescribed treatment for severe OSAS is CPAP, which has a long-term (>12 years) adherence of approximately 70% [23].

Motor vehicle collision rate has been shown to improve in several uncontrolled treatment studies on OSA. In a controlled study, George et al. identified that CPAP treatment had a favorable effect on driving capacity in domestic drivers with OSAS [5]. Using Canadian provincial governmental data, where MVC record-keeping is mandatory, the 3-year MVC rate fell in OSAS patients treated with CPAP from 0.18 (pretreatment) to 0.06 (post-treatment) accidents per annum, a similar level to the normal control population. Untreated OSAS patients continued to have a high MVC rate over the 3-year observation period. However, from a population of 740 OSAS patients, within which driving records were available for 582, CPAP was adhered to in 317 of 353 patients, and 2-year follow-up was available in only 210 drivers. Moreover, the population studied wes not commercial drivers. Thus, generalization of this study to commercial drivers should be made with caution.

Whether to screen for OSA or not is a vexed question. To screen for any disease, the disease must be important, highly prevalent and have an identifiable chronic latent state with an accurate, feasible and cost-effective screening test and a treatment with favorable impact [24]. Can OSAS be classified as such in commercial drivers? We know the ‘disease’ OSA is related to cardiovascular complications, MVCs and sleepiness [25]. A chronic latent state exists, as indicated by only a third of patients with OSA being symptomatic [4]. Although an approximately 10% night-to-night variability in OSA severity exists, a diagnosis can be made accurately and feasibly with a cost-effective screening test (unattended limited channel cardiopulmonary and body position monitoring). Effective treatment is available with lifestyle changes, CPAP, dental devices or surgery. Concerning the financial impact of successful screening in the USA of OSAS in commercial drivers, Gurubhagavatula et al. discussed the financial cost of the MVC which can be attributable to either the driver error or the test, plus the treatment must have a high treatment acceptance (>75%) [26]. A single expensive screening test (i.e., an in-laboratory polysmonogram) would not be a cost-effective screening tool. An alternative approach, argued by the authors, was for a screening questionnaire (MAP), followed by oximetry and selective polysmongraphy to those patients in the intermediate- or high-probability groups. This has the potential to provide a screening tool with high sensitivity and specificity at a significantly lower cost than using polysomnography alone [27].

The American College of Chest Physicians has recently stated that management of driving safety in patients with OSAS who drive commercially should involve individual assessment by a medical examiner, and that no sleep study, questionnaire or similar, will allow an absolute answer [28].

The US federal motor car safety association recently made the following recommendations [Pack A, Pers. Comm.] following assessment of the available data: screening for OSAS with weight, height and questionnaire (20r). If the BMI is greater than 33 kg/m², Berlin questionnaire positive [20] or the driver is judged high-risk by clinical assessment (e.g., diabetic, cardiac condition, previous sleep related or unexplained MVC), he/she should undergo either an in-laboratory polysomnography or portable unattended limited-channel monitoring (high-fidelity oximetry, oronasal airflow, heart rate variability to determine apnea type [29]), and an estimate of sleep–wake time over several days using wrist watch-sized motion sensors (actigraphy).

The ethics of screening for OSAS in the commercial driver is a concern. In contrast to domestic drivers, commercial drivers are held to higher levels of scrutiny because they drive larger-sized vehicles that may carry hazardous goods or multiple passengers for long periods of time over many miles. Whether self-employed or an employee, there is always a strong economic incentive to ‘keep those wheels rolling’.

When a driver perceives their license and employment may be at risk due to the results of the OSAS or any other assessment, they might avoid participating in screening programs or provide false answers to screening questionnaires about sleepiness or sleep apnea symptoms [30]. In practice, response rates in large screening studies have varied from 32 to 72%, well short of 100% of drivers [8,9]!

A successful screening program will need assurance of doctor–patient (driver) confidentiality, and the capacity to approach drivers independently of employees and companies. Screening for OSAS and other sleep disorders must have the support of employee bodies. While a regulated compulsory process (similar to assessment of visual acuity) would ensure participation, it may encourage false answers on subjective assessments or possibly even substitution of a healthy friend or relative for sleep study monitoring if licenses are at risk.

Many commercial drivers are now aged over 60 years (Pack A, Pers. Comm.) and the turnover of drivers is high, thus any screening program will need to be ongoing, repeated periodically for individuals and appropriately funded. Questions of funding a screening process (and treatment) of the driver or the employee, will need debating. In Australia, workplace and transport insurers have participated in funding preliminary screening studies, and this may be a model for further screening programs.

Finally, any screening program for sleepiness and OSA is only as good as the treatment offered for the underlying disorders. A broad program that modifies driver schedules and sleep habits in response to sleep restriction, as well as ensuring optimal treatment of OSAS with ongoing objective monitoring of treatment (e.g., CPAP adherence), would be preferable. Government legislation and policing of driving hours is required in parallel [31]. Should drivers with a positive OSAS screening test who decline treatment be reported to their government driving authority? It is the opinion of these authors that a nonpunative attitude be adopted for most drivers, and consider restriction (or cancellation) of commercial driver licenses only for those who are severely symptomatic of OSA and are not adequately treated.

Expert commentary

Obstructive sleep apnea syndrome is a common condition that is associated with sleepiness as well as cardiovascular disease. Domestic drivers with OSAS have a two- to sevenfold greater risk of MVCs, whereas, in commercial drivers, the risk is small and is outweighed by sleep restriction. Identification of OSAS should be paralleled by identification of sleep restriction. Attitudes to reverse sleep restriction will require education, employer and social change, with greater importance on sleep duration and quality. This may require an attitudinal change within the transport industry that places sleep on par with diet and exercise as important factors contributing to a healthy lifestyle and greater productivity, with fewer MVCs and, thus, cost to the community. Treatment of OSAS requires long-term follow-up to ensure adherence to therapy.

Five-year view

Screening for reversible or treatable conditions will be mandatory for all commercial drivers. This screening program will include sleep apnea and sleep restriction via a simplified portable un­attended sleep study (e.g., high-fidelity oximetry, heart rate variability and body position) plus actigraphy. Drivers will be assessed for fitness for work upon presentation for work and during shifts. The avoidance of sleep-related MVCs will also require the industry to better educate drivers about sleepiness and develop more intelligent motor vehicles with drowsiness sensors that will alert the driver and be recorded in a ‘black box’ available for an ‘automotive autopsy’ following a serious MVC.

Financial & competing interests disclosure

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

No writing assistance was utilized in the production of this manuscript.