Integrity of Forensic Science Hits Another Pothole

Today, most American law enforcement jurisdictions impose an absolute limit on 0.08% of blood ethanol as the cut point for impaired driving. The gold standard blood alcohol measured by headspace analysis. This was a little complicated to perform well so various firms have developed and sold instruments that are less reliable, and probably less costly. However, investigative technical reporters are finding that the instruments and indefensible laboratory practices have combined to give forensic science another black eye.

A front-page article in the NY Times provides some statistics and detailed case histories. These show that the problem with breath ethanol measurement is large and widespread. About 1 million people in the USA are convicted of driving under the influence (DUI) of each year. Often the only supporting evidence is results from an analysis of a suspect's breath using one of a series of instruments that have evolved over the last 50 years.

Cited problems include:

  • Lack of calibration
  • Improper reference (calibration) standards
  • Lack of internal standards for the assay
  • Poor maintenance
  • Lack of expertise of the operators
  • Operational safeguards are ignored during testing
  • Programing errors in the software
  • Ignoring concerns from technical experts during pre-purchase evaluations
  • Using an outdated computer operating systems and programs
  • Lack of transparency in the operating and data generation programs
  • Fraud in certification of installation qualification

Blood ethanol concentration (BAC) usually involves headspace gas chromatography. However, drawing a blood sample often requires a warrant, which takes time to obtain. Plus, there is the issue of what happens when the draw station involves a long delay time for the suspect. Ethanol is rapidly metabolized. Law enforcement feels the need for a quick and low-cost assay protocol, ideally field deployable. Several firms have designed, and marketed instruments designed to meet this need.

The misuse of calibration standards is cited as a weak point. If standards are old or not properly maintained, the ethanol will preferentially evaporate compared to water. The operator can recalibrate by adjusting the amplification of the signal to output the label value of the standard. This will also excessively amplify the values from the real samples producing a high result.

The story goes on to discuss serious errors in laboratory management including falsifying reports. The net result has been that 42,000 breath tests in Massachusetts and New Jersey have been thrown out as evidence in DUI prosecutions. The impact in other states will certainly add to this subtotal.

The authors have uncovered and documented a major failing in the American forensic science system. I hope that this becomes a major topic for NIST’s Organization of Scientific Area Committees for Forensic Science (OSAC). This is the decade-old program to restore a science base to forensic science. It has come a long way, but not far enough. Further, I think that this is a great opportunity for a national proficiency testing program with real samples.