Results of a Randomized Trial of Web-based Responsible
Beverage
g Service Training:
g WayToServe
WayToServe.org
y
org
g
G. Woodall
G
Woodall, University of New Mexico Center on Alcoholism,
Alcoholism Substance Abuse,
Abuse and Addictions
((CASAA);
); R. Saltz,, Prevention Research Center,, Pacific Institute for Research and Evaluation,,
Berkeley; D.
D Buller,
Buller Klein-Buendel,
Klein Buendel Inc.,
Inc Golden,
Golden CO; R.
R Starling,
Starling University of New Mexico CASAA;
and P. Stanghetta, Paula Stanghetta Associates, Kitchener, ON, Canada
,1752'8&7,21
Research indicates that alcohol use and misuse is the third
leading cause of preventable death in the United States
(M kd d ett al.,
(Mokdad
l 2004).
2004) The
Th National
N ti
l Highway
Hi h
T ffi Safety
Traffic
S f t
Administration ((NHTSA,, 2010)) found that fatalities due to
drunk driving (BAC .08+)
08+) accounted for approximately 32% of
all traffic deaths in 2009.
2009 On average,
average a person is killed in an
alcohol related driving
g crash everyy 50 minutes in the United
States Research indicates that up to half of people arrested
States.
for driving-while-intoxicated
driving while intoxicated (DWI) were drinking in alcohol
serving
i establishments
bli h
(
(e.g.
b
bars,
restaurants, lounges)
l
) prior
i
to their arrest (Ventura County Behavioral Health, 2007).
Recent research from Naimi,
Naimi et al.
al (2009) indicated that over
10 percentt off binge
bi
di k
drinkers
(fi or more drinks
(five
d i k on a single
i l
occasion)) drove a vehicle within 2 hours of their most recent
binge drinking occasion,
occasion and that drinking in licensed
establishments
t bli h
t accounted
t d for
f
over 54% off those
th
di i
driving
episodes.
p
Current research in Responsible
p
Beverage
g Service
Training (RBS) suggests that such training can be an effective
means of preventing over-service
over service of alcohol,
alcohol and ultimately
can reduce the incidence of drunk driving;
g e.g.,
g alcohol
servers may help prevent alcohol
alcohol-impaired
impaired crashes by
discouraging over-consumption
over consumption of alcohol among their
patrons..
t
R
Recent
t reviews
i
off the
th RBS literature
lit t
suggestt that
th t
training methods and implementation are an important factor
in determining whether RBS training can achieve desired
alcohol-related
l h l l t d outcomes.
t
Th
There
are now a number
b
off
commerciallyy available web-based RBS training
g courses
(e g learn2serve.com,
(e.g.,
learn2serve com tipsalcohol.com);
tipsalcohol com); however,
however to date,
date no
systematic examination of the effectiveness of these online
RBS courses has been p
published. Reported
p
here is the first
randomized trial effectiveness evidence for a web
web-based
based RBS
on premise (liquor by the drink) training – WayToServe.org,
on-premise
WayToServe org
an RBS training
t i i website
b it systematically
t
ti ll developed
d
l
d and
d tested
t t d
via a grant from NIH/NIAAA (R01 AA014982) .
0(7+2'
• Th
The New
N
M i
Mexico
Al h l and
Alcohol
d Gaming
G i
Di i i
Division
(AGD)
provided
p
o ded a list
st o
of a
alcohol
co o se
serving
g estab
establishments
s e ts from
o which
c
one hundred fifty-five (N=155) were then randomly
selected
l t d across four
f
communities
iti
i
in
N
New
M i
Mexico.
Establishments had to have at least a beer and wine
license and serve alcohol independent of any food order.
order
• These 155 on-site
on site premises were then randomly assigned
t either
to
ith the
th WayToServe
W T S
t i i
training
( 77) or to
(n=77)
t the
th usuall
and
a
d cus
customary
o a y training
a
g ((n=78)
8) p
provided
o ded in live
e c
classes
asses in
New Mexico by local trainers.
trainers Establishments randomized
t receive
to
i the
th WayToServe
W T S
t i i were then
training
th contacted
t t d by
b
project
p
j
staff for p
purposes
p
of recruiting
g them into the study.
y
• Individuals of at least 21 years of age were recruited and
trained to act as “pseudo
pseudo-patron
patron” (PP) buyer and observer
t
teams.
Th teams
The
t
consisted
i t d off an individual
i di id l who
h attempted
tt
t d
to p
purchase alcohol while acting
g obviouslyy intoxicated at
the alcohol serving establishments and an observer of the
purchase
h
attempt
tt
t who
h acted
t d as an innocent
i
t bystander.
b t d The
Th
use of p
pseudo-patrons
p
in p
previous studies to evaluate the
effectiveness of various alcohol server training models is
well established (Gliksman et al.,
al 1993; McKnight,
McKnight 1989,
1989
1991, 1993; Russ & Geller, 1987; Saltz, 1989).
)
• The buyer/observer
y
teams were blind to the experimental
p
condition of each premise.
premise
• Design:
D i
Th design
The
d i employed
l
d in
i this
thi investigation
i
ti ti was a 2
((level of training:
g WayToServe
y
vs. Usual and Customaryy –
UC - training) x 4(time of assessment: baseline vs.
vs
immediate post-training
post training vs.
vs 6 month follow-up
follow up vs.
vs one year
follow-up)
p) mixed factorial design.
g
&HQWHURQ$OFRKROLVP6XEVWDQFH$EXVHDQG$GGLFWLRQV
&HQWHU
RQ $OFRKROLVP 6XEVWDQFH $EXVH DQG $GGLFWLRQV &$6$$
8QLYHUVLW\ RI 1HZ 0H[LFR
8QLYHUVLW\RI1HZ0H[LFR
KWWSFDVDDXQPHGX
S
•Premises
P
i
were assessed
d att baseline,
b
li
i
immediate
di t postt
5(68/76
training,
g, 6 month p
post-training,
g, and one-year
y
post-training
p
g
intervals Pseudo Patron (PP) assessments were employed
intervals.
A 2 (level of training) x 4 (time of assessment) repeated
to assess whether apparently intoxicated patrons would be
measures analysis
ly i off variance
i
was conducted
d
d on the
h data,
d
served in p
premises at each of the four time assessment
and found significant main effects for training (F(1,
points At each assessment point,
points.
point the PP/observer teams
264)=5 55 p=.019),
264)=5.55,
p= 019) time of assessment (F(3,
(F(3 792)=34.07,
792)=34 07
would visit each establishment twice.
twice The purpose of
p=.0001)
0001) and
d a significant
i ifi
t training
t i i
b time
by
ti
i t
interaction
ti
d i g two visits
i i
was to reduce
d
the
h possibility
ibili y off
conducting
((F(3,792)=2.88,
(3, 9 ) 88, p
p=.035).
035) Planned
a ed t-tests
es s (o
(one-tailed)
e a ed) a
at eac
each
collecting anomalous data per a single visit to an
assessment period found no significant differences for refusal
establishment All assessment occurred between summer
establishment.
rate
t between
b t
W T S
WayToServe
and
d UC training
t i i
att baseline
b
li
2009 and
d fall
f ll 2011.
2011
((WayToServe
y
refusal rate mean=36%;; UC refusal rate
•Although
g a number of dependent
p
measures were collected,,
mean=35%)
mean
35%), but a highly significant difference at immediate
we shall focus here on rate of refusal: how often apparently
post training (t(288)=3.265,
post-training
(t(288)=3 265 p<.00001)
p< 00001) between WayToServe
intoxicated PPs were refused service in any given
((refusal rate mean=68%)) and UC ((refusal rate mean=49%).
)
establishment.
bli h
Thi is
This
i the
h primary
i
outcome off the
h
At six months follow
follow-up
up, no differences were detected
investigation.
between WayToServe (mean refusal rate=69%) and UC
t i i (mean
training
(
refusal
f
l rate=65%).
t 65%) However,
H
att one year postt
The WayToServe website is based on the content of the
assessment, a significant difference was detected between
state mandated New Mexico Alcohol Server Training
WayToServe training (mean refusal rate=68%) and UC
curriculum. WayToServe
y
was developed
p
through
g several
t i i
training
(
(mean
refusal
f
l rate=58%),
t 58%) t(272)=1.739,
t(272) 1 739 p=.0025).
0025)
focus groups across New Mexico with members of the
Figure
g
1 displays
p y these means.
target population in order to determine proper terminology,
terminology
idea groupings,
id
i
and
d sensitive
iti topics.
t i
O
Once
th initial
the
i iti l website
b it
was developed, project members conducted additional
multimedia focus groups with alcohol servers and RBS
policy
li makers
k
(i l di the
(including
th director
di t off AGD) throughout
th
h t the
th
state in order to refine the design
g and structure of the earlyy
beta version of the website.
website Prior to conducting the efficacy
t i l WayToServe
trial,
W T S
was subjected
bj t d to
t a beta-test
b t t t in
i four
f
communities in New Mexico. The final version of the website
consists of six multimedia,
multimedia interactive content modules that
include information on the following:
1)Alcohol
)
laws - What one can do legally
g y as a server, what
one must do as a server, and the penalties one could face
for failure to uphold the alcohol laws.
laws
2)P
2)Preventing
ti
service
i
t minors
to
i
- Information
I f
ti
i presented
is
t d
that will help in identifying underage patrons, including the
finer points of checking an ID and information on strategies
used
d by
b minors
i
i attempting
in
tt
ti to
t obtain
bt i alcohol.
l h l
3)Alcohol
)
and p
pregnancy
g
y - Information about Fetal Alcohol
Spectrum Disorders (FASD) is presented,
presented and includes
frequently asked questions about FASD.
FASD
4)Facts
)
about alcohol - Information is p
presented about the
effects of alcohol and how to recognize a patron who has
had too much to drink.
drink Also,
Also information is provided
regarding
di
th definition
the
d fi iti
off a standard
t d d drink
d i k and
d the
th
definition of blood alcohol concentration levels.
5)The intoxicated patron - Here,
Here the user learns about the
active
ti role
l he
h or she
h plays
l
i refusing
in
f i further
f th alcohol
l h l service
i
to intoxicated p
patrons and p
preventing
g p
patrons from driving
g
under the influence of alcohol.
alcohol It includes a skill building
activity for these purposes.
purposes
6)Responsible
)
p
alcohol sales p
practices - In this section,, the
user will have the opportunity to learn about the SERVE
system and also to apply and practice each one of its
system,
points.
i
Th SERVE system is
The
i a summary off what
h has
h been
b
presented in the previous five sections. Below are screen
captures from WayToServe.
WayToServe
Fig. 1
80.00%
80
00%
70.00%
67 61%
67.61%
69.57%
65.07%
68 70%
68.70%
60.00%
58 33%
58.33%
50.00%
40.00%
49.67%
Intervention
Premises
Control
P
Premises
i
35 71%
35.71%
34.84%
30.00%
20.00%
10.00%
0.00%
B
Baseline
li
P t
Post
Intervention
6 Month
M th
Follow up
Follow-up
O Y
One
Year
Follow up
Follow-up
Overall,
O
ll these
th
results
lt indicate
i di t that
th t WayToServe
W T S
t i i led
training
l d
to significantly
g
y more refusals to apparently
pp
y intoxicated
customers than the Usual and Customary training available
in New Mexico.
Mexico Although refusals also increased secularly
in the UC training
gg
group,
p significant
g
refusal rate differences
were detected in 2 of the three follow
follow-up
up assessment
points most notably at One Year follow-up.
points,
follow up
6800$5<
The most prominent
Th
i
finding
fi di g in
i the
h present investigation
i
ig i iis
that web
web-based
based training (provided by WayToServe)
provides superior refusal rate outcomes when compared to
d customary
t
lilive ttraining.
i i
Gi
Given th
the llack
k off any
usuall and
a do
ed trial
a research
esea c o
on o
other
e web-based
eb based RBS
S
randomized
programs this is an important addition to the RBS research
programs,
lit t
literature.
M
Mostt notably,
t bl th
the iimpactt off W
WayToServe
T S
y
follow-up
pp
period,, a finding
g
maintained over the year
suggesting that the longitudinal impact of this training
persists It is further worthwhile to note that the level of
persists.
refusal ((close to 70% at Year One follow-up)
p) represents
p
a
high level of refusal compared to other similar outcomes in
the literature
literature.
5()(5(1&(6
Gliksman, L
Gliksman
L., McKenzie
McKenzie, D
D., Single
Single, E
E., Douglas
Douglas, R
R., Brunet
Brunet, S
S., & Moffatt
Moffatt, K
K. (1993)
(1993). The role
of alcohol providers in prevention: An evaluation of a server intervention programme
programme.
Addiction 88,
Addiction,
88 1195
1195-1203
1203.
g , A. J. ((1989).
) Development
p
and Field Test of a Responsible
p
Alcohol Service
McKnight,
Program. Volume 3. Final results prevention program. (DOT HS 807449). Washington D.C.:
National Highway Traffic Safety Administration.
McKnight, A. J. (1991). Factors influencing the effectiveness of server-intervention education.
J
Journal
l off Studies
St di on Al
Alcohol,
h l 52
52, 589
589-397.
397
McKnight A
McKnight,
A. JJ. (1993).
(1993) Server intervention
intervention. Alcohol,
Alcohol Health
Health, & Research World
World, 17
17, 76
76-83.
83
Mokdad A
Mokdad,
A.H.,
H Marks,
Marks JJ.S.,
S Stroup
Stroup, D
D.F.,
F and Gerberding
Gerberding, JJ.L.
L (2004)
(2004). Actual causes of death
in
the United States,, 2000. JAMA,, 291,, 1238-1245.
Naimi, T.X., Nelson, D.E., and Brewer, R.D. (2009). Driving after binge drinking. American
Journal of Preventive Medicine, 37, 314-320.
National Highway Traffic Safety Administration (2010). Alcohol-impaired driving. Retrived June
30 2011 ffrom http://www.nhtsa.gov/staticfiles/ncsa/pdf/2010/811385.pdf
30,
htt //
ht
/ t ti fil /
/ df/2010/811385 df
Russ N
Russ,
N. W
W., & Geller
Geller, E
E. S
S. (1987)
(1987). Training bar personnel to prevent drunken driving: A field
evaluation American Journal of Public Health,
evaluation.
Health 77
77, 952-954
952 954.
Saltz, R. F. (1989). Research needs opportunities in server intervention programs. Health
Education Quarterly,
y 16, 429-438.
Ventura Country Behavioral Health, Alcohol and Drug Prevention Division (2007). Ventura
County Place of last Drink Survey. Ventura, CA: author
Poster presented at the 25th CASAA Anniversary Celebration at the Hotel
Andaluz Albuquerque
Andaluz,
Albuquerque, NM on November 77, 2014