Case Study - The Personal Breath Test Device

Part 2: When an Alcohol Monitor Gets It Wrong

From - Counterpoint Volume 3: Issue 3 - Article 2 (December 2018)

An article for participants in the myCAMprogram

Jan Semenoff, BA, EMA
​Forensic Criminalist

Article information:

900 words (approximately 4-5 minutes)

Why This Case Matters to You

This case involved a Personal Alcohol Tester — a portable breath testing device often used in court-ordered alcohol monitoring. The person wearing it was required to test to prove they were not drinking alcohol, as part of their supervised visitation with their child.
Over the monitoring period, they had more than 100 negative tests. But there were two “positive” results — and both were disputed. The first was proven to be a false positive caused by mouth alcohol (residue from toothpaste with alcohol in it). This article focuses on the second positive reading — and why the evidence suggested the device itself had malfunctioned.
If you are ever accused of drinking by your monitoring device, this case shows why it’s important to:
  • Know how your device works.
  • Keep your own records and photos of any unusual events.
  • Understand the kinds of errors that can happen.

​How Personal Breath Test Devices Work

Portable Breath Testers have three main parts:
  1. A fuel cell ethanol sensor that measures alcohol in your breath.
  2. A built-in camera with facial recognition to confirm the right person is taking the test.
  3. Cellular transmission to send results to a central database, where they can be shared with courts, probation officers, or other approved parties.
Because they use a fuel cell, these devices do not have a “slope detector” — a feature that can help identify mouth alcohol contamination. This means they cannot tell if alcohol came from a drink or from another source (like mouthwash). Most manufacturer’s guidelines list common products that can trigger false positives, but it’s up to the user to avoid them.

The Second Positive Reading – What Happened?

Here’s what the device data showed:
  • Test #155 – Taken at 3:47 PM on December 14, 2018 – Result: 0.000 (zero).
  • More than 8 hours later, Test #157 – Taken at 11:50 PM – Result: 0.022 g/dL.
​At first glance, it looked like the person might have consumed alcohol in that time. But there were several serious problems in the data.

​Discussion of event 2

​A breath test (#155) was obtained at 1547 hours, 14 December 2018 with a reported reading of 0.000 (zero) grams. Slightly more than 8 hours elapsed before the next breath test (#157) was obtained at 2350 hours, 14 December 2018 with a reported value of 0.022 grams/dL. But, the device was programmed to obtain a random sample about every 4-hours. Something was clearly wrong with the data presented, especially considering the evidence regarding the operation of the device obtained by the test subject himself. 
Error #1 – Wrong “Missed Test” Report
The printed report claimed there was a missed test at 8:00 PM (#156), but it showed no device ID. In reality, the person did attempt a test at 7:43 PM, and they even photographed the Soberlink screen showing an “E” for error. That photo didn’t match what the official report said.
Error #2 – Missing Scheduled Tests
The PBT is supposed to take a new sample every 30 minutes after a missed test until a reading is recorded. That didn’t happen. Instead, it went back to a 4-hour schedule — something it wasn’t supposed to do. This points to a possible software malfunction.
​
Error #3 – Incorrect Confirmatory Test Timing
After a positive reading at 11:50 PM, the unit should have taken a confirmation test after 15 minutes. Instead, it waited 26 minutes. If the user had missed that second test, the device should have kept asking every 30 minutes — but it didn’t follow that protocol until much later.
After the missed test at 0031 hours (#159) the unit returned to 30-minute intervals.
​We can speculate on a variety of causes as to the reliability or unreliability of the numerical results at 2350 hours (#157) and 0016 Hours (#158). First, we have to recognize that they may reflect true BrAC values. Equally, they may be the result of the device being in some sort of undetermined error condition. We simply cannot know for certain, but we do know they are reported at the end of more than four-hours of cumulative errors. More likely than not, a simple re-boot of the device might have cleared the error (think of your tablet device, cellphone, or laptop freezing up... What is the first thing you do?)

Could the Reading Have Been Accurate?

We can’t completely rule it out. There are three possibilities:
  1. ​Drinking Soon After the 3:47 PM Test – To reach 0.022 by 11:50 PM, the person would have had to drink over 9 standard drinks in that time.
  2. Drinking Just Before the 11:50 PM Test – This would have meant having about 1 standard drink right before the test — highly unlikely since the person knew they were about to be tested.
  3. A Combination of Both – Drinking throughout the evening, totalling between 1 and 9 drinks.
​But, the person firmly denied drinking at all — and had clear photo evidence of a device error.
The photo evidence tipped the event in his favor. Without it, he would have lost his visitation rights - and face sanctions from the Family Court for violating his conditions.
Using the same assumptions that we used in the previous article when we discussed the first event:
Recall that the test subject was an adult male who weighed about 220 pounds and stands 6’2” in height. A male of that height and weight will receive the BAC equivalent of 0.017 g/dL of blood for each Standard Drink consumed. We have also discussed Standard Drinks - a Standard Drink is measured as follows:
  • 12 ounces of beer at 5% alcohol content
  • 5 ounces of wine at 12% alcohol content, and
  • 1.5 ounces of hard spirit liquor at 40% alcohol content.
   

Why the Malfunction Theory Makes Sense

The device:
  • Didn’t follow its own testing schedule.
  • Didn’t accurately record or report its own error messages.
  • Showed time and numbering errors in the log.
​Any one of these would raise questions about reliability. All three together strongly suggest a malfunction.

​The reliability of breath alcohol readings

​As we discussed in Part 1, reliability in any forensic measurement is established by creating specific acts, practices and conditions under which the testing is performed. If any one of these is compromised or performed under sub-standard circumstances, the resulting reported value is also sub-standard, and therefore must be considered unreliable. This is very much a “digital” condition. A reading is either considered reliable or unreliable. There is no middle ground. 

​It should again be mentioned that the technology employed in the PBT device, a fuel cell, is not robust enough to include a means of detecting mouth alcohol contamination. A wait period of 20 minutes prior to sample provision is required.
 
What happened with these two reported readings? 
​

What the Court Decided

The judge reviewed the technical evidence, the photo of the device error, and the fact that the person had over 100 negative tests. The court ruled:
  • The “positive” readings could not be relied on.
  • The device had malfunctioned.
  • The device must be replaced at no cost to the person.
  • The person was allowed supervised visits over the holidays.

Key Lessons for You as a Participant:

  • Document Everything – Take photos or screenshots if your device gives an error message.
  • Know Your Device Rules – If it’s supposed to retest after 15 minutes or keep testing every 30 minutes, make sure it does.
  • Compare Reports to Reality – If the official report doesn’t match what actually happened, that’s important evidence.
  • Question Sudden Positives – Especially if they happen after device errors or missed scheduled tests.
  • Push for Replacement – If there’s clear evidence your device isn’t working properly, request a replacement immediately.

​Practice Tip:

A reading is either reliable or it isn’t — there is no “kind of” reliable. If the process is broken at any step, the reading’s credibility is in doubt. Always keep your own backup records. Use your camera or video to document what happened. Protect your progress...

    Question? Comments? Start the conversation here:

Send me your questions or comments:
​
Comments and questions will be posted here with their responses:

For further study:

  1. Dubowski, K.M., Acceptable Practices for Evidential Breath-Alcohol Testing, Center for Studies of Law in Action, Borkenstein Course Materials, Indiana University, May 2008.
  2. Dubowski, K.M., Quality Assurance in Breath-Alcohol Analysis, Journal of Analytical Toxicology, Vol. 18, Oct 1994.
  3. Gullberg, R. G., Breath Alcohol Measurement Variability Associated with Different Instrumentation and Protocols, Forensic Science International 131 (2003) 30-35.
  4. Jones, A. W., Concerning Accuracy and Precision of Breath-Alcohol Measurements, Clinical Chemistry, 33/10, 1701-1706 (1987).
  5. Jones, A. W., Evidence-Based Survey of the Elimination Rates of Ethanol from Blood with Applications in Forensic Casework, Forensic Science International 200, 1-20 (2010).
  6. Sterling, Kari, The Rate of Dissipation of Mouth Alcohol in Alcohol Positive Subjects, The Journal of Forensic Science, 2011.