Dynamic Assessment and Referral System - Evaluation

Screening process. All patients who verbally consent to being screened by the RA will
undergo a computer assisted screening for substance abuse (Rapid Screener) administered by
the research assistant. Patients who use tobacco, misuse alcohol, or use illicit drugs will
be further screened for eligibility to participate in the randomized trial. Eligible
patients will be invited to participate and will sign the consent form. Participants will
then be assigned to one of the two conditions by the computer using a random number
generator. Results obtained as part of the Rapid Screener will be for research purposes
only. They will not be reported to healthcare providers, unless the participant is enrolled
into the study and is assigned to the DARSSA Intervention condition (see below).

Minimal Intervention Control. All patients that give consent to participate in the study
("participants") who are randomly assigned to the control condition will complete the
computerized DARSSA for assessment purposes only. The reports will not be printed or
dynamic referrals generated, and all patients will receive treatment-as-usual by their ED
providers. This will preserve the treatment-as-usual nature of the baseline while
maintaining consistency in the administration of the DARSSA between phases of the study.
Although ED clinicians will know that a study is being conducted in the ED, they will not be
informed of the results of the research assessment, nor will they be given any instruction
or training beforehand to change their standard screening, counseling, and referral
practices. This means that substance use screening is likely to be inconsistent and
counseling or referrals by provider rare. To maintain ethical standards, the RAs will
provide a substance abuse treatment resource list (passive referral list) to all subjects
who screen positive, even though this often does not happen during routine clinical
practice. Additionally, suicidal ideation is not directly assessed as part of the study,
but it is possible that participants will spontaneously report suicidal ideation. If this
happens, the RA will notify the treating physician. Both of these efforts could conceivably
introduce interventions the patient may not otherwise have received, but there is no viable
alternative if the investigators are to ensure patient safety.

DARSSA Intervention. All participants randomized to the DARSSA Intervention will be given
instructions for how to complete the assessment. Since the DARSSA is self-administered and,
like all Polaris products, does not require computer literacy, instructions will be simple
and do not require a clinician to administer. Once completed, the treating emergency
physician will be expected to (1) give substance using patients the Patient Feedback Report,
(2) recommend they review it carefully, and (3) encourage them to consider following up with
the referrals. Any additional assessment or counseling will be optional and left to
individual provider practice. We considered creating a standardized counseling protocol but
decided against it. We want to maintain an ecologically valid evaluation. In the "real
world," providers would not be given extensive health behavior counseling training or asked
to follow a standardized counseling protocol when implementing the DARSSA. Training
providers on health behavior counseling would act as an additional intervention in its own
right, and it could potentially mask our ability to determine what effect the DARSSA alone
is having.