Week 2 Lecture.pdf

CVEN4405 Human Factors in Civil and Transport Engineering Week 2 Lecture Recreational substances & driver behaviour Dr Milad Haghani Research Centre for Integrated Transport Innovation (rCITI) School of Civil and Environmental Engineering Recreational substances & driver behaviour • • • • • • • • • • • • Types of recreational substances Impact of alcohol & marijuana on central nervous system Alcohol and relative risk of accidents Epidemiological studies Experimental studies Meta-analyses on the effect of alcohol The effect of drugs Challenges of establishing the effect of drugs Cannabis and relative risk of road accidents Publication bias Experimental evidence on the effect of cannabis Prescription drugs Types of recreational substances and their impact on the central nervous system • Alcohol o Alcohol is a depressant that slows down CNS activity. o It inhibits the release of neurotransmitters like glutamate, which leads to impaired cognitive function, reduced inhibitions,and relaxation. o Alcohol can impair motor skills, coordination, and reaction time, affecting tasks such as driving. o https://www.youtube.com/watch?v=B-EmeQg40wE • Cannabis (Marijuana) o Cannabis contains compounds like THC that interact with cannabinoid receptors in the brain. o Effects may include altered perception, impaired memory, and changes in mood. o It can also affect coordination, attention, and reaction time, making it unsafe for driving. o https://www.youtube.com/watch?v=ISUXrjBXHsE • Stimulants (e.g., Cocaine, Methamphetamine) • Opioids (e.g., Heroin, Prescription Painkillers) • Hallucinogens (e.g., LSD, Psilocybin Mushrooms) • Prescription Medications (e.g., Benzodiazepines, Sleep Aids) • Designer Drugs (e.g., MDMA, Synthetic Cannabinoids) Alcohol impaired driving Alcohol impaired driving • Drinking alcohol impairs a wide range of skills necessary for performing driving tasks including alertness, divided attention, vigilance, visual tracking, and quick reaction time to everchanging information and the ability to execute maneuvers. • Blood Alcohol Concentration (BAC), expressed as the percentage of alcohol in deciliters of blood, helps provide an indication of how much alcohol an individual has consumed in the past few hours. • A positive BAC is present in 33-69% of fatally injured drivers and in 8-29% of nonfatally injured drivers. • At a BAC of 0.05, a driver suffers impairment in eye movements, glare resistance, visual perception, reaction time, steering ability, information processing, and other aspects of psychomotor performance. Alcohol impaired driving Relative risk of a crash by BAC with covariates and without covariates. Source: Blomberg, R.D., Peck, R.C., Moskowitz, H., Burns, M., Fiorentino, D., 2005. Crash risk of alcohol involved driving: a case-control study, final report to the National Highway Traffic Safety Administration. Dunlap and Associates, Inc., Stamford, CT. Blood Alcohol Concentration (BAC) Drink driving campaigns https://www.nhtsa.gov/campaign/drive-sober-or-get-pulled-over https://www.nhtsa.gov/campaign/ride-sober Alcohol impaired driving Total and alcohol-related traffic fatality rates in the United States per 100 million vehicle miles traveled (VMT), 1977e2004. Source: National Institute on Alcohol Abuse and Alcoholism (2006b). Alcohol impaired driving https://www.alcoholproblemsandsolutions.org/biphasic-curve-shows-how-alcohol-affects-us/ The limb of the BAC curve Breslin, F.C., Mayward, M. and Baum, A., 1994. Effect of stress on perceived intoxication and the blood alcohol curve in men and women. Health Psychology, 13(6), p.479. Alcohol impaired driving • The risk of a motor vehicle crash increases as BAC increases and is complicated further with increased demands of the driving task. • The risk of a single vehicle fatal crash for drivers with a BAC of 0.10-0.14 is 48 times higher than the risk for sober drivers. With BACs of 0.15 or higher, the risk is 382 times higher. • Approximately 85% of BAC-positive drivers involved in crashes are above of 0.08 (Yi, Chen, & Williams, 2006), and more than 50% of drivers in fatal crashes have BACs at or above 0.16 Alcohol impaired driving Alcohol impaired driving Peck, R.C., Gebers, M.A., Voas, R.B. and Romano, E., 2008. The relationship between blood alcohol concentration (BAC), age, and crash risk. Journal of safety research, 39(3), pp.311-319. Alcohol impaired driving • For young, inexperienced drivers under the influence of alcohol, the risk is even greater. • Inexperienced drivers with BACs of 0.05 have 2.5 times the risk of a crash compared to more experienced drivers. • Behavioural tolerance is thought to explain this increased risk. The repeated performance of a task in association with alcohol consumption can lead to the development of an adaptation referred to as learned or behavioural tolerance. That is, for a well-learned task such as repeatedly driving a certain route home from a bar after drinking, behavioral tolerance decreases the impairment ordinarily associated with alcohol consumption. However, when conditions unexpectedly change (e.g., an animal dashing in front of one’s car), the behavioural tolerance effect is negated. Risk Factors Related to Alcohol-Impaired Driving Demographic characteristics associated with increased risk of driving under the influence include: • Being a male (males are approximately twice as likely to drive after drinking compared to females) • Being younger than 35 years old • Being unmarried • Having a high school education or more • Being employed • Being born in a single-parent family • Living in a rural setting • Being a smoker • Having a prior DUI conviction • Having a family history of alcohol abuse … Policies related to Alcohol-Impaired Driving • A general review of studies of the effect of lowering the BAC limit in Australia, Europe, and the United States was conducted in 2006 and concluded that lowering the BAC limit to .05 would be likely to reduce alcohol-related crashes in the United States. • A study in 1994 found that lowering the BAC limit from .08 to .05 g/dL in New South Wales, Australia, significantly reduced fatal crashes on Saturdays by 13%. • There is a strong case for setting a .05 BAC limit for driving. Laboratory and test track research shows that the vast majority of drivers, even experienced drinkers who typically reach BACs of .15 or greater, are impaired at .05 BAC and higher with regard to critical driving tasks • A 2017 meta-analysis of international studies on lowering the BAC limit in general found an 11.1% decline in fatal alcoholrelated crashes from lowering the BAC to .05 or lower. Policies and regulations related to Alcohol-Impaired Driving • By 1988, all states in the United States had enacted minimum legal drinking age (MLDA) laws making it illegal for those younger than 21 to purchase or possess alcohol. This provided a basis for implementing a zero BAC limit for drivers aged 20 and younger (generally defined as a BAC of .02 or greater). • These zero-tolerance laws for youth have proven effective in reducing fatal crashes involving underage drinking drivers. A meta-analysis conducted in 2001 of the studies of zero-tolerance laws found reductions of 9%–24% in fatal crashes. Experimental evidence on Alcohol-Impaired Driving Acute alcohol consumption during “simulated” driving • The impact of acute alcohol consumption on simulated driving performance • Control (‘no alcohol’ or ‘placebo alcohol’ ingestion) vs acute alcohol ingestion • Primary measures: • • Standard deviation of lane position (SDLP) [lateral vehicle control measure] • Standard deviation of speed (SDSP) [longitudinal vehicle control measure] 50 repeated-measures trials (n = 962 participants) derived from 17 original publications were reviewed Experimental evidence on Alcohol-Impaired Driving Forest plot displaying the effect of acute alcohol intake (BAC range: 23–100 mg·dL−1) on difference in mean SDLP Irwin, C., Iudakhina, E., Desbrow, B. and McCartney, D., 2017. Effects of acute alcohol consumption on measures of simulated driving: A systematic review and meta-analysis. Accident Analysis & Prevention, 102, pp.248-266. Experimental evidence on Alcohol-Impaired Driving Forest plot displaying the effect of acute alcohol intake (BAC range: 23– 100 mg·dL−1) on difference in mean SDSP Irwin, C., Iudakhina, E., Desbrow, B. and McCartney, D., 2017. Effects of acute alcohol consumption on measures of simulated driving: A systematic review and meta-analysis. Accident Analysis & Prevention, 102, pp.248-266. Experimental evidence on Alcohol-Impaired Driving Forest plot displaying the effect of acute alcohol intake (BAC range: 62-78 mg·dL−1) on difference in mean number of Lane crossings (LC) Irwin, C., Iudakhina, E., Desbrow, B. and McCartney, D., 2017. Effects of acute alcohol consumption on measures of simulated driving: A systematic review and meta-analysis. Accident Analysis & Prevention, 102, pp.248-266. Experimental evidence on Alcohol-Impaired Driving Correlation between change in BAC level (mg·dL−1) and change in SDLP Irwin, C., Iudakhina, E., Desbrow, B. and McCartney, D., 2017. Effects of acute alcohol consumption on measures of simulated driving: A systematic review and meta-analysis. Accident Analysis & Prevention, 102, pp.248-266. • • Alcohol consumption significantly increased standard deviation of lane position (SDLP) by 4.0 ± 0.5 cm Alcohol consumption significantly increased standard deviation of speed (SDSP) by 0.38 ± 0.10 km·h−1 Recreational substances & driver behaviour • • • • • • • • • • • • Types of recreational substances Impact of alcohol & marijuana on central nervous system Alcohol and relative risk of accidents Epidemiological studies Experimental studies Meta-analyses on the effect of alcohol The effect of drugs Challenges of establishing the effect of drugs Cannabis and relative risk of road accidents Publication bias Experimental evidence on the effect of cannabis Prescription drugs Cannabis and its relation to road accidents Drugs and its relation to road accidents • Everybody knows that consuming alcohol before driving increases the risk of being involved in a road accident. The more you drink, the higher the risk. Does this also apply to drugs? • One can easily think that the use of some drugs would increase accident risk, for example, drugs that make you sleepy, produce hallucinations, or otherwise influence cognitive functions. • It has, however, proved considerably more difficult to estimate the relationship between drug use and accident risk, than to estimate the corresponding relationship for alcohol. Challenges in establishing this relationship 1. The determination of the dose taken of a drug and the level of impairment caused by it while driving. 2. Controlling for other factors influencing accident risk in addition to the drug of primary interest. 3. How best to summarise the relationship between drug use and accident involvement, that is, either as a point estimate or as a functional relationship. 4. How to test and adjust for the potential presence of publication bias in studies of driver drug use and accident involvement. 5. How to assess study quality and its relationship to estimate the risk of accident involvement associated with drug use. Challenges in establishing this relationship 1. Case control studies: A case-control study is a type of observational research design commonly used in epidemiology and medical research to investigate the causes of a particular condition or disease. In a case-control study, researchers compare individuals who have a specific condition (cases) with individuals who do not have the condition (controls). The goal is to identify factors or exposures that may be associated with the development of the condition or disease. --- A case sample, usually injured drivers treated at hospitals, is compared to a control sample, often drivers stopped in roadside surveys. In this study design, it may be challenging to collect high quality data on the dose of drugs present in the body for both case and control drivers. Blood samples provide the best basis for determining the dose. Taking a blood sample is an invasive procedure, not easily administered roadside to drivers, most of whom are likely to be negative for drugs. To circumvent this difficulty, some studies have used less reliable sources of data to determine the dose of drugs, such as samples of saliva or urine, prescribed dose and self-reported use. 2. Culpability studies: Drivers involved in accidents and judged to be at fault are compared to drivers involved in accidents judged not to be at fault. If both groups of drivers are recruited at medical facilities, it may be easier to collect high-quality data on drug use than in roadside surveys. Challenges in establishing this relationship • The risk of becoming involved in a road accident depends on a host of factors. Drug use is just one of them. In order to correctly estimate the contribution to risk attributable to drug use, a study should, ideally speaking, control for all other factors that are known to influence accident risk. In practice, that is not possible. • Easily observable factors, such as age and gender are controlled for in many studies. • A minority of studies control for annual driving distance. An unbiased comparison requires that drivers using drugs should drive the same annual distance as drivers not taking drugs. • The absence of control for driving distance in most studies of the risk of accident involvement associated with drugs is a source of uncertainty in the results. Publication bias and study quality Study quality is a concept that does not have a standard definition and that escapes precise measurement (Greenland, 1994). However, nearly all researchers agree that studies vary in quality and that findings may be related to study quality. Can we place equal trust in the findings of a study relying on self-reports only and not controlling for any confounding factors, as in a well-controlled study containing precise data on the dose taken of a drug that was still active during driving? Most researchers would say “no.” The relationship between quality score, on a scale ranging from 0 to 1, and estimates of the risk associated with the use of antidepressant drugs. *The odds ratio is a way to measure the likelihood of an event happening in one group compared to another group. Publication bias • Publication bias denotes a tendency not to publish studies if the findings are not statistically significant or go in the opposite direction of what researchers expected and are therefore regarded as difficult to interpret or explain. • Publication bias is a type of bias that occurs in research when the publication of research findings is influenced by the nature and direction of the results. It typically happens when studies with positive or statistically significant results are more likely to be published, while studies with negative or non-significant results are less likely to be published. This can skew the overall body of evidence, leading to a distorted view of the true effect of a particular intervention or relationship. • Publication bias occurs when research studies are selectively published based on the direction or statistical significance of their results. In other words, studies with results that align with a particular hypothesis are more likely to be published, while studies with opposing or non-significant results may remain unpublished or are less likely to be accessible. • Publication bias can lead to an overestimation of the true effect size in a meta-analysis or systematic review. Funnel plot • A funnel plot is a graphical representation used in meta-analysis, systematic reviews, and other forms of evidence synthesis. It is designed to visually detect potential publication bias in a collection of studies on a particular topic. The plot typically consists of a funnel-shaped scatterplot, with each study represented as a point on the graph. • The y-axis of the funnel plot usually represents a measure of study precision (e.g., sample size or standard error), while the xaxis represents the effect size or outcome measure (e.g., the treatment effect). • In the absence of publication bias, smaller studies with lower precision may scatter more widely around the average effect size, while larger studies with higher precision will cluster closer to the average. • In an ideal world, if there were no publication bias, we would expect the points on the funnel plot to be evenly distributed, forming a symmetric funnel shape. • One of the key signs of publication bias is funnel plot asymmetry. If the funnel plot appears lopsided, with a gap or clustering of studies on one side, it suggests potential publication bias. Publication bias Publication bias Publication bias and study quality Hypothetical example of funnel plots. “A” Asymmetric funnel plot with publication bias. “B” Symmetric funnel plot without publication bias. OR: odds ratio Publication bias and study quality Example of symmetrical funnel plot. The outer dashed lines indicate the triangular region within which 95% of studies are expected to lie in the absence of both biases and heterogeneity (fixed effect summary log odds ratio±1.96×standard error of summary log odds ratio). The solid vertical line corresponds to no intervention effect Make sure you do not confuse “forest plots” with “funnel plots” Experimental evidence on the effect of marijuana driving Moderating influence of driving experience and task demand – Compared the effects of three doses of cannabis and alcohol: placebo, low and high – Both alone and in combination – Compared performance of young, novice drivers and more experienced drivers Lenné, M.G., Dietze, P.M., Triggs, T.J., Walmsley, S., Murphy, B. and Redman, J.R., 2010. The effects of cannabis and alcohol on simulated arterial driving: influences of driving experience and task demand. Accident Analysis & Prevention, 42(3), pp.859-866. Experimental evidence on the effect of marijuana driving Both cannabis and alcohol were associated with increases in speed and lateral position variability The effect of cannabis dose and driving experience on the standard deviation of steering angle (panel A) and mean headway (panel B). Prescription drugs Studies were included if they: 1) used an epidemiologic design ensuring that exposure (prescription opioids) preceded the outcome 2) included exposure to intravenous, oral or transdermal prescription opioids 3) had an appropriate comparison group that was not exposed to opioids or other psychoactive substances 4) presented quantitative data and at least one measure of association Some of the underlying studies of this meta-analysis Prescribed and illicit drugs Prescribed and illicit drugs Concluding remarks • The risk of road accident associated with the use of drugs is poorly known. • It is more likely that the use of drugs increases the risk of accident than the opposite. Causality of the relationship has not been established. • Most studies have no precise information about the dose taken of a drug. Very few studies have been able to test for a dose–response relationship between the dose taken of a drug and the size of the increase in risk. This means that a key criterion for assessing the causality of the statistical associations found is largely unknown. • Most studies do not control for a very important potentially confounding factor, annual driving distance. It is unlikely that regular users of illicit of prescription drugs drive the same distance per year as nonusers of drugs. • Most tests for publication bias suggest that there is publication bias. If the results of these tests are taken at face value, they imply that estimates of risk are considerably exaggerated for many drugs. Thus, to give some examples, the estimate of risk adjusted for publication bias was 2.74 for the risk of fatal accident associated with use of amphetamine; the crude estimate was 5.70. Recreational substances & driver behaviour • • • • • • • • • • • • Types of recreational substances Impact of alcohol & marijuana on central nervous system Alcohol and relative risk of accidents Epidemiological studies Experimental studies Meta-analyses on the effect of alcohol The effect of drugs Challenges of establishing the effect of drugs Cannabis and relative risk of road accidents Publication bias Experimental evidence on the effect of cannabis Prescription drugs

Week 2 Lecture.pdf

Related

Transcript

Tags

Upgrade to continue