Diagnosing alcoholism in high-risk drinking drivers: comparing different diagnostic procedures with estimated prevalence of hazardous alcohol use.

In several European countries, drivers under influence (DUI), suspected of an alcohol use disorder (AUD, 'alcoholism') are referred for diagnostic examination. The accuracy of diagnostic procedures used in diagnosing AUD in the DUI population is unknown. The aim of this study was to compare three prevalence estimates of AUD based on a structured clinical interview (SCID), a restrictive diagnostic procedure (RDP) and usual clinical diagnostic procedure (CDP), with a prevalence estimate based on sensitivity and specificity data of biological markers of excessive use of alcohol in non-judicial samples. The latter unbiased estimate provides an external yardstick against which the biased patient-based prevalence estimates in this special sample can be evaluated. The unbiased estimate derived from sensitivity and specificity data resulted in a prevalence estimate of excessive use of alcohol between 74 and 82%, which is much higher than the three diagnostic procedures. SCID identified maximally 5% of alcoholics found with the unbiased estimate. RDP identified > or =31% of the unbiased estimate, while CDP identified > or =60% of the unbiased estimate. The high chance of false positive diagnosis, however, makes CDP unacceptable in the legal context of AUD diagnosis in DUI populations.