Reduced Suspension with Ignition Interlock Conduct Review: A quantitative impact evaluation of an Ontario policy-based intervention

Author(s): Tracey Ma, Patrick Byrne, Yoassry Elzohairy

Slidedeck Presentation Only (no paper submitted):

5A - Ma Byrne Elzohairy



Ontario introduced the Reduced Suspension with Ignition Interlock Conduct Review (CR) program in August 2010. Before this date, drivers convicted of an alcohol-impaired driving offense under the Criminal Code of Canada were subject to a license suspension period followed by an ignition interlock (II) condition period. During the latter period, the driver was required either to refrain from driving, or to only drive a vehicle with an II device installed. The CR program is a modification of the pre-existing program, in which eligible participants can have their license suspension period shortened upon mandatory installation of an II device. The CR program was put into place to increase the effectiveness of the pre-existing program in deterring offenders from committing subsequent acts of drinking and driving. An extensive literature review and an evaluation of the pre-existing program have already been completed.


To evaluate the impact of the CR program on recidivism, participants in this intervention will be compared to those in the pre-existing program to determine whether between-group differences exist in a) the incidence of driving violations committed during the license suspension period; and b) the probability of committing a subsequent alcohol-impaired driving offense both during and after the II condition . Furthermore, c) any additional deterrent effect that may have resulted from extended use the II device will be evaluated. II device extensions are required of drivers who have, in the last 3 months of their II condition, failed any number of breath tests required to either start the engine or to keep the engine running.


The three research questions described above will be addressed through a matched cohort study. To examine driving violations during suspension, the cumulative incidences between exposure groups (CR program vs. pre-existing program) will be compared. To examine re-offense during and after II condition, the probability of re-offense will be modelled as a function of the exposure and elapsed time. This type of probability modelling will also be used to examine whether II extensions have an additional deterrent effect, where exposure groups are defined by the number of interlock extensions. This type of categorical exposure will allow us to infer a potential dose-response relationship.

Analyses are currently underway.


Previous research has examined the deterrent effects of II devices both during and after device installation. The majority of these studies utilize either logistic regression or Cox regression to examine re-offense odds and survival time, respectively. However, both methods have shortcomings. This research uses a different approach to model probability per unit time, thus encompassing the strengths of both regression approaches without encountering their shortcomings. In addition to this new method, our work is novel in that it examines a dose-response relationship that may result from policy-mandated II extensions. Evaluation of such policy intervention has yet to be conducted in the Ontario context.


The Reduced Suspension with Ignition Interlock Conduct Review policy intervention is being evaluated for its specific deterrent effects.