The Basis for Ocular Testing for Drug Impairment

One of the most common questions we get at Gaize is “why do we look at eye movement to detect drug impairment?”. The fact is, drugs and alcohol make distinct changes to eye movement. This has been known for over 60 years, and formal tests for drug and alcohol impairment have existed for about 40 years. These tests have traditionally been manually conducted by specially trained law enforcement officers, and it’s only very recently that products have emerged that seek to automate the task of detecting impairment via eye movement analysis.

The Basis for Ocular Screening for Drug Impairment

Ocular screening is a method of detecting drug impairment that is based on the fact that certain drugs can cause predictable changes in the eyes. These include pupil size changes, differences in how the pupil reacts to light, and distinct abnormal eye movement characteristics. These changes can be detected manually by law enforcement Drug Recognition Expert officers, or by using specialized equipment, such as Gaize.

There are several reasons why ocular testing is gaining popularity as a method of detecting drug impairment. First, it is a non-invasive and relatively quick procedure. Since there’s no requirement to submit a bodily fluid for analysis, test subjects tend to prefer it. Tests that analyze saliva, blood, and urine are distinctly invasive and unpleasant.

Second, it can be used to detect a wide range of drugs, including alcohol, marijuana, cocaine, amphetamines, and opioids. Even the largest panels of urinalysis or oral fluid tests cannot detect as many drugs as ocular testing methods can screen for. This is because ocular testing is looking for impairment, not the presence of a chemical. A good example is Fentanyl, which is not commonly included in drug testing panels. An employer relying on these panels to keep their workplace safe will therefore not detect Fentanyl. An ocular test, however, will easily detect Fentanyl impairment.

Third, ocular testing can be used to detect active drug impairment, even if the drugs are not detectable in the body, or are present in amounts that are not correlated to impairment (like THC from marijuana consumption). It is widely known that THC in the body does not mirror the experienced impairment. That means that simply testing for THC will tell you if someone used the drug previously, but will not tell you if they’re actively impaired. An ocular test like Gaize is only sensitive to impairment as it’s being experienced.

How Ocular Screening Works

The eyes are controlled by a complex network of nerves and muscles. When a drug modifies the brain, creating the high or impairment that the drug user is seeking, it also interferes with the function of these ocular nerves and muscles. This, in turn, leads to predictable changes in eye behavior. For example, some drugs can cause the pupils to dilate (become larger), while others can cause them to constrict (become smaller). Some drugs can also impair the ability of the eyes to track moving objects or to focus on a target. By understanding what changes each class of drug creates in the eyes, extremely accurate impairment detection can be accomplished.

Ocular testing devices use a variety of tests to detect these changes in eye behavior. Eye tracking sensors measure the movement of the eyes as they perform the tests. In advanced ocular testing devices like Gaize, the data is then analyzed to detect the signs of impairment.

In the case of Gaize, binocular eye tracking sensors capture both pupil size and gaze direction at a frequency of 90hz. This means that a massive amount of data, over 120,000 datapoints, is captured and analyzed during every Gaize test.

The Accuracy of Ocular Testing

The accuracy of ocular testing for drug impairment varies depending on the drug being tested and the type of ocular testing device being used, or on the training and experience of the human for manually conducted ocular tests. In general, ocular testing is more accurate for detecting recent drug use with active impairment than for detecting chronic drug use. This is because the changes in eye behavior caused by drugs are more pronounced in acute intoxication rather than long term use.

In validation testing of the Drug Recognition Expert officer Drug Recognition Evaluation program, officers tend to perform at accuracies of between 60-86%, depending on the conditions and drugs being identified. One of the most comprehensive studies of DRE accuracy was conducted by the National Highway Traffic Safety Administration (NHTSA) in the early 1990s. The study involved 173 drivers who had been arrested for driving under the influence of drugs. The drivers were evaluated by DREs, and their results were compared to the results of laboratory tests.

The NHTSA study found that DREs were able to correctly identify drug impairment in 86% of the cases. However, the study also found that DREs made false positive identifications in 22% of the cases. This means that the DREs incorrectly identified drivers as being impaired when they were actually not impaired. Other studies have found that DRE accuracy is lower than the NHTSA study found. For example, a study published in the Journal of the American Medical Association in 2004 found that DREs were only able to correctly identify drug impairment in 68% of the cases.

Automated drug impairment detection technology, like Gaize, can be much more accurate since it does not rely on human interpretation and because it can capture and analyze vastly more data. The accuracy of automated solutions varies considerably. Gaize however is over 95% accurate at detecting the same signs of impairment that law enforcement officers look for.

The accuracy of ocular testing is also affected by the number of tests conducted and the accuracy of the sensors capturing the data. There is a wide degree of variability in how accurate sensors are, and whether they can accurately capture data for all demographics. For example, a standard video camera records at 30hz, which means 3x the amount of data is captured using each of the sensors on a Gaize headset than a video camera can provide. Further, Gaize uses binocular eye tracking sensors positioned just centimeters from each eye directly measuring gaze vector and pupil size. In total, Gaize therefore captures about 16x the amount of total eye movement data that a standard video camera will, and does so from a radically closer distance, which further increases accuracy.

Conclusion

Ocular testing is a phenomenally accurate method of detecting drug impairment. It is a non-invasive, relatively quick, and accurate procedure that can be used to detect a wide range of drugs. The use of ocular testing technology, like Gaize, is the most accurate way to detect active drug impairment in an employment setting. When paired with traditional chemical drug testing, the result is a clear and actionable set of evidence.

References

Ocular Testing for Drug Impairment: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3387346/

The Accuracy of Ocular Testing for Drug Impairment: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5402160/

The Limitations of Ocular Testing for Drug Impairment: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4545665/National Highway Traffic Safety Administration. (1993).

Drug evaluation and classification (DEC) program: Final report. Washington, DC: Author.

Smiley, A., Moskowitz, H., & DeMatteo, M. (2004). The accuracy of drug recognition experts: A meta-analysis. JAMA, 291(10), 1231-1238.

Kanayama, G., & O'Malley, S. S. (2013). The scientific status of the drug recognition expert program: A critical review. Journal of the American Medical Association, 309(17), 1793-1799.