Common DUI Defense Challenges in Forensic Alcohol Analysis Rod G. Gullberg Washington State Patrol Seattle, Washington
DUI Litigation Over 1.4 million DUI arrests annually in US (FBI,UCR) More than other felonies employing forensic evidence - homicide, rape, robbery, aggravated assaults, burglary, larceny-theft, auto theft, arson DUI affects broad cross-section of society DUI cases connect forensic sciences with legal system more than any other single offense More significant legal challenge
Legal Context of the Forensic Sciences Forensic toxicology is a mandated science Scientific evidence is persuasive and compelling Adversarial system/knowledge incomplete Evidence subject to interpretation/challenges expected DUI is a very costly offense Legal challenges do not have to be sensible (scientifically or legally) Challenges differ depending on: pre-trial/trial Many are re-statements of challenges heard for years Occasionally some are very creative Be objective learn from issues Can result in program improvements
DUI Litigation Washington State Approx. 42,000 DUI arrests in 2002 Approx. 96% breath alcohol 4% blood alcohol/drug Approx. 250 discovery requests per month Approx. 60 court appearances per month Forensic toxicologists 80% of court appearances are breath cases 20% of court appearances are blood cases
Alcohol Test Protocol Washington State Breath Alcohol BAC Datamaster Duplicate breath samples (± 10% of mean) Internal standard and external simulator standard 15 minute observation Annual Quality Assurance Procedures Data collection, printout of results Blood Alcohol Headspace gas chromatography Duplicate analyses - different instruments/columns Control standards every 10 samples Daily calibration
Common DUI Defense Challenges
Challenge 1 Uncertainty Near Critical Per Se Levels Reasonable doubt due to per se nature of law Multiple critical levels make this a common issue - 0.02 for minors - 0.04 for commercial drivers - 0.08 DUI - 0.15 enhanced penalties Can depend on whether criminal or civil case - Beyond a reasonable doubt - Preponderance of the evidence Issue often framed inappropriately - 0.01 margin of error
Response Uncertainty Near Critical Per Se Levels Must acknowledge measurement uncertainty Be prepared to compute confidence interval X ± t α / 1 2, df = SD n Need duplicate analyses (other jurisdictions) Discuss with prosecutor Provide documentation of computations Automatically perform subtraction factor - European countries
Challenge 2 Biased Results Due to Interfering Substances Acetone historically the principle concern Workplace exposure to VOC s more common now Research showing over 100 VOC in human breath Breath instruments at 0.01 g/210l ethanol equivalent An issue included for cases near per se limits An issue because statutes define offense as alcohol
Response Biased Results Due to Interfering Substances Acetone long recognized as a potential interferent Breath instruments at 0.01 g/210l ethanol equivalent Breath instruments with dual technology Officer s testimony important regarding exposure Acetone necessary to yield 0.01 g/210l equivalent Offer a deal in select cases
Challenge 3 Vacutainer Preservative/Anticoagulant There was no material in vacutainer The materials are not correct in nature/proportion Contamination or removal of materials Expired date on vacutainer
Response Vacutainer Preservative/Anticoagulant Blood received without clotting Certificates available from manufacturer Gray top vials defined by FDA regarding contents Tampering impossible since still had vacuum Officer can note that observed powder material Careful wording of administrative rules
Challenge 4 Inaccuracy of Simulator Thermometers temperature must be 34 0 ±0.2 When adding uncertainty of ± 0.1 results incorrect Administrative rules are foundational Check of thermometers found approx. 25% incorrect
Response Inaccuracy of Simulator Thermometers temperature must be 34 0 ± 0.3 as measured by a certified thermometer Careful wording of administrative rules Have details in policy/procedure manuals Annual certification of thermometers Careful record keeping Does not affect breath alcohol analysis Effect is immeasurable (0.5 0 error! 3.2% error) Database monitor instrument performance over time Thermometer testing probably exceeds most protocols
Challenge 5 Traceability of Simulator Thermometers using a reference thermometer traceable to standards maintained by NIST Issues regarding meaning of traceability Issues regarding meaning of standards Protocols not those approved by NIST Pre-trail suppression hearings
Response Traceability of Simulator Thermometers Careful wording of administrative rules Elimination of this language from administrative rules Submit digital thermometers directly to lab below NIST Careful record keeping Have program authority testify to legislative intent
Challenge 6 Storage of Simulator Solutions in Plastic Analytical chemistry literature recommends glass - for low level VOC - trace analysis, environmental chemistry May explain the depletion of simulator results Results in improperly calibrated instrument
Response Storage of Simulator Solutions in Plastic Example of irrelevant comparison Employ appropriate laboratory practice - sealing and labeling bottles - employ 500 ml polyethylene bottles - avoid using if below ¼ in gallon bottles - label with date, limit shelf life to one year Re-analyze bottles kept over different periods of time Demonstrate stability Alcohol at much higher concentrations than trace work Evaluate effect of calibrating instrument Practice customary to the discipline
Challenge 7 Repair and Error Record History Repair and certification records are kept routinely Provided through discovery or public disclosure Used to argue that client s test must be biased Database records will record errors Errors on one instrument compared to others
Response Repair and Error Record History Keep careful documentation Provide full disclosure of all materials testify to Explain reasons for maintaining detailed records Possible Web site development - Washington -South Carolina Argue independence of each test Instrument design to abort test if improper Not every attempted test should be acceptable (5%) Thorough explanation of printout document elements Database could allow development of control charts - statistical control
Challenge 8 Computation of Widmark Estimates Widmark s Equation: N Wr = 0.8 [ Ct + βt] [ fl.oz. / drink] C t = N [ 0.8][ fl.oz. / drink] Wr βt Used to show that defendant s two beers yields zero Used to show the large number of drinks necessary If introduced by prosecution then discredited
Response Computation of Widmark Estimates Large body of literature Compute the equation in both ways - results are functions of uncertain variables - input given by defendant ( two beers )? - input given by the instrument/protocol results? Explain impact of using BrAC to estimate BAC Explain the large number of drinks necessary Explain average BrAC of those arrested for DUI
Challenge 9 Blood Sampling Following IV/Transfusion IV contaminated sample regarding alcohol analysis Transfused blood may have contained drugs Documentation regarding times, arms not adequate
Response Blood Sampling Following IV/Transfusion Careful documentation of time, location, arm, etc. Fluid IV should dilute sample and bias alcohol low Some research on this Obtain sample of transfused blood for analysis
Challenge 10 Instrument Software Inadequate Computerized instruments require software code Begin by demanding copies of software code Defense experts evaluate and provide opinion Defense experts may or may not evaluate instruments Argue that errors exist in code that bias results Argue for suppression if do not obtain code
Response Instrument Software Inadequate Advise manufacturer of potential challenge/request Request that software be independently tested Emphasize the software as one of several components Sound protocol/quality control ensures fit-for-purpose Certification testing ensures fitness-for-purpose Errors in code may simply be different style Approve specific versions in administrative rules Print software version on printout Software experts from other disciplines relevance?
Challenge 11 Results Do Not Reflect Time of Driving Sampling in blood and breath cases always delayed Changes in concentration time curve BrAC (g/210l) 0.15 0.1 0.05 0 0 5 10 15 Time (hrs)
Response Results Do Not Reflect Time of Driving Two-hour Laws Research suggesting drunk drivers descending BrAC Experimental research showing rapid onset of peak Large BrAC measurement variability
Challenge 12 Records Generated By Other Personnel Often a need to testify to work done by another - instrument calibration/certifications - simulator solution preparations/analysis - repair/maintenance forms Argue document not admissible through another Notes of other scientist not admissible (State v Nation) Division of work in Tox Lab - one analyzes sample for marijuana - another analyzes for cocaine - extractions take time
Response Records Generated By Other Personnel Regular Business Record Exception to hearsay rule Personnel should be custodian of the records Job descriptions - custodian of the records May depend on whether pre-trial or trial proceeding Be able to explain procedures performed Records part of program policy manuals May need to provide copies of records to others Re-organize work effort in lab
Challenge 13 Inconsistency Between Reports and BrAC Officer s written report carefully scrutinized Observations and SFST not reflective of BrAC/BAC Supplemented with Widmark estimates - showing large number of drinks estimated - how can person not do poorly on SFST Subpoena of other DUI reports - show repetitive nature - suggest that officer just changes the name Generally these are issues going to the weight
Response Inconsistency Between Reports and BrAC Carefully written reports - careful to detail - reports different and unique Careful compliance with NHTSA protocol for SFST Forensic computation of Widmark estimates Research regarding alcohol tolerance Drinking labs with careful Widmark estimates
Miscellaneous Challenges Tongue jewelry biasing breath alcohol result Previous error record requires full re-certification Swab material used prior to blood draw Compliance with observation time periods Defective Miranda/implied consent warnings Switching of samples in gas chromatography (Georgia) Incomplete discovery provided BAC agreement requirement of 0.01% Inadequate warranty on instruments Bias due to radio frequency interference Endogenously produced alcohol (gut fermentation) Interferent threshold must be subtracted from result Drinking after the accident ( hip flask defense ) Variation between replicate results Variation in sequential standard results Dynamics of pulmonary physiology
Principles For Case Preparation Pre-trial hearings vs. jury trial Full disclosure of all requested evidence to defense Thorough discussion with prosecutors Openness in discussions with defense Acknowledge uncertainties
Suggested References 1. Jones, A.W., Top Ten Defense Challenges Among Drinking Drivers in Sweden, Med. Sci. Law, Vol.31 No.3, 1991, pp. 229-238 2. Pollack, S., Observations on the Adversary System and the Role of the Forensic Scientist: Scientific Truth v. Legal Truth, JFS, Vol.18 No.3, July 1973, pp. 173-177. 3. South Carolina s Web Site: http://www.sled.state.sc.us/default.htm 4. Washington s Web Site: http://www.wa.gov/wsp/reports/flbindex.htm#breath 5. NIST Web Site: http://www.nist.gov/traceability/nist%traceability%20policy-external.htm