An Adaptive Intervention for p Reducing College Student Alcohol R - - PowerPoint PPT Presentation

An Adaptive Intervention for p Reducing College Student Alcohol‐ R l t d C Related Consequences:

A SMART Pilot Study Proposal

Megan Patrick & Daniel Almirall SRC, ISR, University of Michigan

Center for the Study of Health and Risk Behaviors

• Univ. of Washington—March 27, 2012
• Univ. of Washington

March 27, 2012

The slide I don’t need at C‐SHRB. The slide I don t need at C SHRB.

• College alcohol and associated negative

College alcohol and associated negative consequences are clear public health problems

• Heavy drinking at age 18 has long term impact
• Heavy drinking at age 18 has long‐term impact,

including symptoms of AUD in adulthood S i i b d

• Scarce intervention resources must be used as

wisely as possible to address college student l h l alcohol use

Another slide I don’t need at C‐SHRB.

• Brief motivational interventions are efficacious
• Our thinking is this
• Our thinking is this…

– Web‐based versions of BMI, which are suitable as universal interventions are especially resource universal interventions, are especially resource‐ efficient and have shown promising effects. In person BMIs (e g BASICS) which due to their – In‐person BMIs (e.g., BASICS), which due to their intensity may be more suitable as indicated interventions, have a strong evidence base. , g

One way to address heavy drinking while i i conserving resources is to

• first utilize universal interventions,
• identify students in need of additional

y treatment,

• and then move these students to a more

and then move these students to a more intensive, indicated intervention. This package of sequenced treatments (universal f ll d b i di t d) i f f d ti followed by indicated) is a form of adaptive intervention.

One (Somewhat Simple) Example of an Adaptive Intervention an Adaptive Intervention

@ Week 6 Baseline: Week 0 > Week 6 Web‐based BMI: delivered one week prior to Binge‐ drinker In‐person BASICS week prior to beginning of Fall semester Non Binge‐ drinkers Watchful Waiting Binge drinking at Week 6 is defined as 4+/5+ drinks from women/men in 2 or more settings over the past 6 weeks. Web-based BMI is a 5 component (intended drinking, expected consequences, social norms, goal-directed motivations, protective behavioral strategies) personalized feedback intervention based on responses to the baseline survey. p p y In-person BASICS (2 session program) will be delivered by a University Health Services professional. BASICS is freely available at the Univ. of Michigan.

Some key issues—useful for building ti l d ti i t ti an optimal adaptive intervention— remain untested in the science of brief alcohol interventions.

• The optimal timing of universal BMI: Do we

intervene with a web‐based BMI prior to or after intervene with a web based BMI prior to or after the beginning of Fall semester?

• Among students who continue to binge drink
• Among students who continue to binge‐drink

despite the universal treatment, what is the added value of a BASICS over a web based BMI Booster? value of a BASICS over a web‐based BMI Booster?

SEQUENTIAL MULTIPLE ASSIGNMENT RANDOMIZED TRIAL (SMART) Study Design to develop an adaptive intervention to prevent and treat college drinking

Measures and

EARLY BMI: S

Binge‐ drinkers

BASICS PROGRAM R A

@ Week 6

Outcomes: @ Weekly: drinking quantity and

Web‐based intervention delivered prior to the Fall semester WEB‐BASED BMI BOOSTERS

Non Binge‐

WATCHFUL WAITING B C

@ Week 0: 1 Week prior to Fall semester

quantity and duration on each day of the previous week, along with a

SAMPLE: College R

Non Binge‐ drinkers

WATCHFUL WAITING C

checklist of consequences and their evaluations. @ Weeks 0 6 12:

College Freshmen DELAYED BMI

Binge‐ drinkers

BASICS PROGRAM R D

@ Weeks 0, 6, 12: Intentions for drinking, perceived norms, motivations,

DELAYED BMI: Web‐based intervention delivered after Fall semester begins

N Bi

WEB‐BASED BMI BOOSTERS E

expected and eval.

• f conseq., protect.
• behav. strategies,

health service

begins

Non Binge‐ drinkers

WATCHFUL WAITING F

health service utilization, GPA Web-based BMI Boosters includes content that is tailored using updated information.

Specific Aims of the SMART

• Aim 1 (Timing Question): To determine the maximally effective

timeframe to deliver the universal, web‐based BMI. ( d ) d h h b

• Aim 2 (Best Second Stage Treatment): To determine whether web‐

based boosters versus BASICS is the best treatment among participants who exhibit binge‐drinking despite the initial web‐ p p g g p based BMI.

• Aim 3 (Best Adaptive Intervention): To determine which is the

best of the 4 pre‐specified adaptive interventions.

• Aim 4 (Tailoring): To determine (i) the types of individuals (on the

basis of baseline measures e g age race intentions to rush) basis of baseline measures, e.g., age, race, intentions to rush) more or less likely to benefit from early vs delayed BMI delivery, (ii) the types of binge‐drinkers (on the basis of baseline measures, and outcomes to initial treatment) more or less likely to benefit from web‐based boosters vs BASICS.

The 4 Embedded Adaptive Interventions Any One Participant May Be Assigned To

EARLY BMI:

Binge

WEB BASED BMI ADAPTIVE INTERVENTION #1 ADAPTIVE INTERVENTION #2 EARLY BMI: Web‐based intervention delivered prior to the Fall semester

Binge‐ drinkers Non Binge‐

BASICS PROGRAM WATCHFUL WAITING EARLY BMI: Web‐based intervention delivered prior to the Fall semester

Binge‐ drinkers Non Binge‐

WATCHFUL WAITING WEB‐BASED BMI BOOSTERS semester

g drinkers

WATCHFUL WAITING semester

Non Binge‐ drinkers

WATCHFUL WAITING ADAPTIVE INTERVENTION #3 ADAPTIVE INTERVENTION #4 DELAYED BMI: Web‐based intervention delivered after Fall semester

Binge‐ drinkers

BASICS PROGRAM DELAYED BMI: Web‐based intervention delivered after Fall semester

Binge‐ drinkers

WEB‐BASED BMI BOOSTERS Fall semester begins

Non Binge‐ drinkers

WATCHFUL WAITING Fall semester begins

Non Binge‐ drinkers

WATCHFUL WAITING

Specific Aims of the SMART Pilot

• Aim 1: To develop and refine the adaptive interventions to

b l t d i th SMART I l d d l i d be evaluated in the SMART. Includes developing and refining components of the initial web‐based BMI and boosters and developing the protocol for linking students boosters, and developing the protocol for linking students between the initial web‐based BMI to the in‐person BASICS program. p g

• Aim 2: To conduct a pilot SMART study to address

acceptability and feasibility concerns about the SMART study design. The pilot SMART will provide (i) provide investigators with the experience to conduct a SMART, and (ii) li i d t f R01 t b b itt d t th (ii) preliminary data for an R01 to be submitted to the NIAAA to fund a larger‐scale SMART.

Why do a SMART pilot first?

Aim 1: Develop and refine intervention modalities Aim 1: Develop and refine intervention modalities

• Focus groups before (show mock‐ups) and after

H d li k bi d i k UM BASICS?

• How do we link binge‐drinkers to UM BASICS?

Aim 2: Feasibility and acceptability

• Can we consent and engage prior to Fall?
• How do we link binge‐drinkers to UM BASICS?

How do we link binge drinkers to UM BASICS?

• How do we classify students (for purposes of 2nd

line txt) who fail to provide Week 6 binge status? line txt) who fail to provide Week 6 binge status?

• Get %age of students who meet binge threshold
• Infrastructure for multiple randomizations
• Can we deliver web‐based BMI boosters via text?

3 questions to the group for discussion

• What else do you anticipate is important for

What else do you anticipate is important for us to think about in the Pilot study?

• Do you have points of concern/confusion
• Do you have points of concern/confusion

about the methodology? D h i f / f i

• Do you have points of concern/confusion

about the scientific justification for this study?

The Interventions Assigned from the Participant’s Point of View.

There are 6 sequences of treatment that are possible. EARLY BMI: Web‐based intervention delivered prior

Binge‐ drinkers

BASICS PROGRAM Txt Sequence A DELAYED BMI: Web‐based intervention delivered after

Binge‐ drinkers

BASICS PROGRAM Txt Sequence D p to the Fall semester Txt Sequence B Fall semester begins Txt Sequence E EARLY BMI: Web‐based intervention delivered prior

Binge‐ drinkers

WEB‐BASED BMI BOOSTERS Txt Sequence B DELAYED BMI: Web‐based intervention delivered after

Binge‐ drinkers

WEB‐BASED BMI BOOSTERS Txt Sequence E to the Fall semester Txt Sequence C Fall semester begins Txt Sequence F EARLY BMI: Web‐based intervention delivered prior

Non Binge‐ drinkers

WATCHFUL WAITING Txt Sequence C DELAYED BMI: Web‐based intervention delivered after

Non Binge‐ drinkers

WATCHFUL WAITING Txt Sequence F to the Fall semester Fall semester begins