Evaluation of the Drug Evaluation and Classification Program: A Critical Review of the Evidence

Abstract

Objective. A critical review of the existing evaluation studies on the Drug Evaluation and Classification (DEC) program was conducted to determine the validity and accuracy of the technique for identifying drivers under the influence of drugs.

Methods. Studies were divided into two categories—laboratory studies and field (i.e., enforcement) studies. A classification process was devised using common criteria based on the toxicology findings (i.e., drug positive or drug negative) and the opinion of the police officer who assessed the driver (i.e., drug positive or drug negative). A series of standard measures (Sensitivity, Specificity, False Alarm Rate, Miss Rate, Corroboration, and Accuracy) were calculated for each to assess the effectiveness of the DEC program.

Results. Laboratory studies do not provide overwhelming support for the accuracy with which officers trained in the DEC program can detect and identify the particular class(es) of drug involved based on psychophysical assessment alone. The detection and identification of the relatively low levels of drugs administered were typically better than chance but many cases were missed. The fact that some drugs were detected with greater accuracy than others suggests that the effects of these substances were more prominently manifested in the symptomology assessed by the DEC procedure. Although field enforcement studies are not as scientifically rigorous as laboratory studies, DEC assessments in an enforcement context have the benefit of information obtained from the arresting officer and from interviews with the suspect. In addition, the drug doses consumed by users are typically much higher than those permitted in controlled laboratory studies. In general, officers trained in the DEC program are able to identify persons under the influence of drugs and to specify the drug class responsible with a degree of accuracy that not only exceeds chance, but in some cases reaches a very high level.

Conclusions. There remains room for improvement in the DEC program. As further research becomes available, either from laboratory or field investigations or both, it needs to be incorporated into the program to enhance its validity and accuracy.

Keywords:

• Drugged Driving
• Drug Evaluation and Classification (DEC)
• Drug Recognition Expert (DRE)
• Impaired Driving
• Impairment Testing

ACKNOWLEDGMENTS

This work was funded by the Royal Canadian Mounted Police.

Initial work on this project was conducted while the senior author was with the Traffic Injury Research Foundation (TIRF). The views expressed in this article are those of the authors and do not necessarily reflect those of their respective institutions or sponsors.

Notes

∗Includes 11 cases where the drug administered was not the drug identified by the DRE. Sensitivity = 144/269 = .535. Specificity = 38/51 = .745. False alarm rate = 13/51 = .255. Miss rate = 125/269 = .465. Corroboration = 144/157 = .917. Accuracy = (144 + 38)/320 = .569.

¹The numbers for individual drugs are based on matches with the first drug indicated by the DRE. These numbers are reported by Shinar and Schechtman (2005) in their Table II (p. 847).

²Accuracy for each drug has been calculated using all 44 matches for the 102 placebo cases.