NHTSA studies Transdermal Electrochemical Alcohol Testing

Experienced Michigan DUI lawyers know the importance of establishing that their clients are alcohol free while on bond or probation.  Many choices are available, and many Michigan DUI lawyers will recommend that their clients attend AA regularly, and courts will often order that the DUI accused undergo testing.  This testing may include the SCRAM anklet.

“SCRAM” is an acronym meaning “secure continuous remote alcohol monitor”.  The SCRAM device, manufactured by AMS, is a tether worn on the offender’s ankle.  AMS claims that the device monitors the use of alcohol by measuring the presence of alcohol as it migrates with perspiration through the offender’s skin.

Michigan DUI Lawyers are consequently often faced with a dilemma – should they recommend SCRAM to their clients?  This is a dilemma because no measuring device is 100 percent accurate 100 percent of the time, and if the SCRAM says a DUI lawyer’s client was drinking when he or she was not, the DUI lawyer will have a tough fight to prove to the court that the device was wrong.

The Science of Transdermal Alcohol Testing:

Much has been written on the science of transdermal electrochemical alcohol testing, and rather than reiterate here what has already been written, several articles are attached to these materials.  Not addressed in these materials, at least in any detail, is the November 2007 NHTSA study entitled “Evaluating Transdermal Alcohol Measuring Devices: Final Report. This study evaluated and compared the two types of currently available sweat-alcohol measuring devices, one manufactured by AMS (Alcohol Monitoring Systems) known as the SCRAM device (Secure Continuous Remote Alcohol Monitor) and the other manufactured by Giner Inc., known as the Wrist Transdermal Alcohol Sensor (WrisTAS™). This study was a comparison of performance in the laboratory and under field conditions.

In summary, some of the findings in the report relative to the SCRAM (only one now commercially available) are as follows:

• The authors were concerned about incidences of false negatives, that is, circumstances under which alcohol was known to have been consumed, but no alcohol alerts were received (devices failed to detect the drinking).
• The devices(s) displayed an inability to read at low levels of alcohol. (less than .02 BAC).
• It was noted that the sensitivity decreases over time (this may have been due to the water accumulation in the devices or because of continued exposure to alcohol, and could be improved in the future).
• A “sophisticated” user of the device can circumvent the technology and “hide” drinking.
• Both units performed more poorly than “expected” accuracy and sensitivity.
• The SCRAM was found to be uncomfortable (especially for women) but was thought to be an “acceptable alternative to jail.”

False Positives:

According to the NHTSA study, “the concept of false positives has two different meanings with transdermal detection: (1) we can say with some confidence that there are few or no events that the devices created that look like drinking but were really not drinking because with drinking logs of well-compensated subjects and BAC results we knew definitively when real drinking occurred, but (2) we also know that there are external sources of ethanol signal that are unrelated to drinking that can cause a transdermal response.”

The study found that:

“We detected one clear TAC event related to shaving cream containing an ethanol product, and there was one event that may have been related to perfume in the environment but we could not definitively identify the source. The shaving cream example did not look like a drinking event since it decayed too rapidly, whereas the other did look like a drinking event and had no clear cause.”

Interesting there were 14 other occasions were the data relative to false positives was literally tossed (disregarded).  In these 14 instances, according to the report, these were instances of self- dosed drinking when no BAC test information was provided by the subjects.

“We know from the drinking logs that these were not genuine false-positive transdermal responses; drinking did occur, but we do not have the corresponding BAC values. Rather than log these as false positives, it was more valid to exclude these from the sample. In the majority of these cases, there was drinking, but the subject did not use the PBT to evaluate the BAC level. Reasons included forgetting to bring the PBT, did not have a breath tube, or as per our instructions, claimed to have had only one drink (the instructions called for logging BAC anytime two or more drinks were consumed in temporal proximity). In a few of those cases, elevated SCRAM™ TACs were an artifact of the initialization process, where voltages may result in artificially high TAC readings until the SCRAM™ ankle bracelet establishes a baseline reading for a non-drinking subject. Actual false positives among the subjects we studied were rare, and when false positives did occur, it was attributable to an undetected external source of alcohol.”

While it would be appropriate to disregard questionable data in a study such as this, it is a bit curious that this data all related to incidences of false positives.  By way of further explanation, or at least, further caution, the NHTSA report indicates:

External sources of alcohol are all around. Many personal hygiene, home, and automotive products contain ethanol, and people who use or are around these products may show positive for exposure to ethanol even though they consumed no alcohol. For example, aerosol Lysol has between 79 and 85 percent ethanol, and dozens of body sprays, such as Avon Naturals, have between 60 and 98 percent ethanol. The National Library of Medicine has a searchable database with product ingredient information (http://householdproducts.nlm.nih.gov). Restricting a search to ethanol-containing ingredients yields 447 consumer products that fall into eight categories, but of these products, 265 (59%) fall into one category: personal care products that are used on or near the skin. The issue of false positives is revisited in a later section dealing with the calculation of ROC curves. The recent paper by Sakai et al. (2006) reported no evidence of false positives in their SCRAM™ research. However, external sources of alcohol can and do result in positive transdermal responses because ethanol or ethanol-like ingredients are in many consumer products. It is important for vendors of this technology to adequately train their staff and customers to appreciate this.

For NHTSA that last sentence is pretty strong language, and in so many words, NHTSA is saying that it important for SCRAM vendors to understand that the possibility of false positives is very real.

Time Delays:

As noted elsewhere, there is a significant time delay between the actual drinking and the expression of the alcohol in the wearer’s perspiration.  This can be as much as four hours (see below).  In this regard the report indicates:

It may be that the delayed peak responses between BAC and TAC reflected residual alcohol laden moisture inside the SCRAM™ unit, leading to a continuing positive alcohol signal and/or a dilution effect of any new alcohol-laden sweat vapor coming in from the skin. This might explain the time delay found with the SCRAM™ units.

This statement is very interesting and helps to explain the sometimes bizarre SCRAM TAC readings, which suggest that the wearer had a very low level of BAC over very long periods of time, such as 12 hours or more.

Discussion:

There is no doubt that the NHTSA study did not give the SCRAM bracelet rave reviews. In addition to the issues identified above, the report indicates:

There were two significant problems with the SCRAM™ bracelets uncovered in this evaluation. These are a higher-than-expected false-negative rate and the possibility that sensitivity and accuracy of the devices decline over time.

If there is such a high incidence of false negatives the DUI lawyer is left to wonder why this device should even be used.  If it is not truly a 24/7 device that will always catch the wearer’s drinking, then why should the court not use the much less expensive (and more “comfortable”) alternatives that exist?

Then there is the problem of the device becoming less effective over time.  This should come as no surprise to experienced DUI lawyers who well know that fuel cell devices fatigue over time.  Here we have an attenuated admission by NHTSA of this fact.

Gender Differences:

As is known to DUI lawyers, breath testing “discriminates” against women.  In this report we find that the SCRAM device also discriminates against women.  The NHTSA report indicates:

On the whole, female TAC accuracy was low, certainly lower than with males, and more female discomfort. A clear false-positive was found during use of “Skintimate” shave cream, a product that contains triethanolamine. These spikes following brief external exposure to alcohols are easily distinguished from elevated BAC (and are not flagged by the automated alert system due to the rapid rise and fall), but there is the larger problem of detecting exposure to various alcohols or industrial chemicals on a regular basis if such substances are in someone’s work environment. AMS is aware of these problems and handles them on a case-by-case basis when they occur.

Here again we have a pretty remarkable statement by NHTSA, which is that workplace exposure can actually cause false positives.  The DUI lawyer should take little comfort in knowing that their clients are protected by AMS taking this issue on a “case-by-case” basis, whatever that may mean.

TAC is NOT BAC:

The journey of alcohol from the stomach to above the skin is variable among individuals, that it is widely acknowledged that sweat-alcohol (transdermal alcohol or TAC) is not the same as blood alcohol (BAC) content.  In fact, TAC can never be thought of as quantitatively accurate.  The NHTSA report puts it this way:

As discussed in more detail elsewhere, the attainable accuracy, however, may only be an approximation of BAC due to subject-specific factors that influence ethanol gas concentration at the skin surface. There is no doubt that the transdermal concept is valid as long as expectations of quantitative parity with BAC are moderated.

Similarly, TAC is not BAC, and the expectation of parity is an impractical expectation to place on this nascent technology. Both interlocks and transdermal sensing need to be judged first on their potential contributions to public safety. Moreover, just as interlock devices have improved in the 20 years since their first adoption, it is reasonable to expect that the transdermal-sensing equipment will also improve. These devices warrant further development and further study.

Final note:

At the end of the report, as a sort of post-script, the NHTSA report has this to say:

In general, the sensitivity and accuracy of these devices were poorer than we expected. But if they are not yet perfected, further product improvement is likely to get them closer. It may be impossible to ever expect the alcohol signal at the skin surface to be a precise estimate of BAC if Anderson and Hlastala (2006) are correct. Their model suggests that the stratum corneum, the outer most layer of skin, and other systemic factors importantly affect the measurable ethanol gas concentration near the skin. Individual differences or state differences within individuals in hydration, temperature, and other factors theoretically affect the transdermal alcohol signal strength greatly.

Get a FREE confidential CASE EVALUATION on your Michigan OWI/OWVI/DUI by calling (248) 306-9159 , or filling out this consultation request form. Call now, there’s no obligation!

Tagged as: dui, SCRAM