DUI risk prediction with alcohol biomarkers, interlock records, and self-report.

The risk of impaired driving can be estimated through the use of several technologies but few programs utilize the full spectrum of predictive technologies and often underutilize the ones they do employ. The record of alcohol ignition interlock breath alcohol (BAC) tests are good predictors of future DUI risk, but still not widely used. These are a good, but imperfect predictor, since drivers other than the DUI offender of record can use the interlock car. Biological markers of alcohol use can more directly target the impairment risk of specific individuals. Use of biomarkers to monitor progress among DUI offenders is growing in Europe but less so in North America. A sample of 537 interlock using DUI offenders from Alberta, Canada have been providing self-report assessments as well as biological specimens at two time points, the beginning of an interlock installation, and typically 6 months later, to determine: a) the predictive relationship between the biomarkers and the record of interlock BAC tests, b) the cross-relationship among the biomarkers, c) the relationship between the biomarkers and self-report assessments, and d) the extent to which the markers, self-report, and interlock BAC tests can correctly predict future (post-interlock) recidivism. Biomarkers under study include conventional alcohol markers such as GGT, ALT, AST, MCV, and percent carbohydrate deficient transferrin (%CDT), but also markers that directly reflect ethanol consumption such phosphatidyl ethanol (PEth), fatty acid ethyl esters (FAEE), and ethyl glucuronide (EtG). Self-report assessments include AUDIT, Drinker Inventory of Consequences (Drinc), the DSM4 Alcohol Dependence assessment from the DIS-C, and Timeline Followback (TLFB) and Temptation Restraint Inventory (TRI). Interlock BAC test data are represented as a proportion of all tests taken that are elevated above .02 g/dL over the full six months as well as for each 2 month interval since installation. Preliminary results show many markers correlate significantly and all in the right direction with interlock BAC tests and self-report indicators. TLFB self-report variables (such as % drinking days, drinks per week) show significant correlations (Spearman and Pearson) with markers and BAC tests. Direct markers, especially PEth, EtG and FAEE, but also GGT, are significantly related to the proportion of elevated interlock BAC tests. Sample size is variable across measures, but continues to grow; post-interlock subsequent DUI events will also grow, but currently with just 9 repeat DUI events there are too few for predictive profiling. Biomarkers are an important monitoring tool and should be used as a supplement to conventional estimates of driver risk. These should be more widely utilized especially in conjunction with alcohol harm reduction or medical interventions.