When refusing alcohol service to a patron the server should do so RBS

Many amateur drinkers are not aware of their alcohol tolerance limit and end up drinking more than their bodies can handle. The same situation arises with alcoholics who cannot control how much they consume. Such patrons can be a liability and risk to themselves and the people around them. Understanding how to carefully deal with over-intoxicated customers is crucial not only for servers that work at alcohol-serving places but for everyone that can come in contact with such patrons.

Here are some practical tips for dealing with an over-intoxicated customer

Tip #1 – Offer Complimentary Beverages: Allow some time to the customer so that alcohol can pass through their bodies. Suggest the patron order a non-alcoholic drink like a mojito, coffee, etc., to allow them some breathing time. If they are not convinced, give them a beverage complimentary on the house. Slowing down consumption time will reduce the effects of alcohol.

Tip #2 – Convince Ordering for Food: Try to convince the group or the patron to order some food before continuing their drink. You can also suggest eating if someone has crossed the standard drinking limits, even if they do not show visible intoxication signs. If they have an empty stomach, alcohol will be absorbed much quicker into the bloodstream.

Tip #3 – Slow Down Serving Drinks: Sometimes, patrons do not accept any food or beverage, even if it is on the house. In such situations, you should slow down your services. Instruct your waiting staff and bartenders to slow the service if a customer request are about alcoholic drinks.

Tip #4 – Talk to the Customer: Intoxicated people seldom listen to others, especially when the talk is about halting the drinks. Nevertheless, you should gently try to explain to them about their consumption and distract them with other options on your menu. If the patron becomes agitated, call your manager to handle the customer. Intoxicated customers, if not controlled, can disrupt other guests, break stuff, and talk mannerlessly to everyone, ruining the establishments reputation. If they understand their actions, you should give them something to eat and call a cab to safely drop take them home. Make sure to document the details of the customer and the cab driver to ensure their safety and record the information in your logbook.

Tip #5 – Refuse Service: If nothing is working to slow down the customer, you should refuse service to them. Do not take their orders or serve any more alcoholic beverages. You can bring them their check so they can pay. If the customer gets belligerent, you can inform your manager or owner and if needed call the police to handle the matter.

About A+ Server Education

Get trained to receive your alcohol certification course by enrolling with A+ Server Education. We have over 15 years of experience and can be trusted to empower you with the knowledge needed to get your permit. If you have any questions about our training courses, reach out to us by calling (503) 740-5509 or our toll-free number (877) 740-5509. You can also mail us for more information.

aCenter on Alcoholism, Substance Abuse, and Addictions – CASAA, The University of New Mexico, Albuquerque, New Mexico

Find articles by W. Gill Woodall

aCenter on Alcoholism, Substance Abuse, and Addictions – CASAA, The University of New Mexico, Albuquerque, New Mexico

Find articles by Randall Starling

bPrevention Research Center – Pacific Institute for Research and Evaluation, Berkeley, California

Find articles by Robert F. Saltz

cKlein Buendel, Inc., Golden, Colorado

Find articles by David B. Buller

dPaula Stanghetta & Associates Inc., Kitchener, Ontario, Canada

Find articles by Paula Stanghetta

aCenter on Alcoholism, Substance Abuse, and Addictions – CASAA, The University of New Mexico, Albuquerque, New Mexico

bPrevention Research Center – Pacific Institute for Research and Evaluation, Berkeley, California

cKlein Buendel, Inc., Golden, Colorado

dPaula Stanghetta & Associates Inc., Kitchener, Ontario, Canada

*Correspondence may be sent to W. Gill Woodall at the Center on Alcoholism, Substance Abuse, and Addictions, The University of New Mexico, 2650 Yale SE, Albuquerque, NM 87106, or via email at: ude.mnu@lladoowg.

Received 2016 Apr 27; Accepted 2018 Mar 25.

Research in Responsible Beverage Service (RBS) training indicates that such training can prevent over-service of alcohol and reduce drunk driving. However, reviews of the RBS literature suggest that the quality of training methods and the stability of implementation are important factors in determining RBS training effectiveness. Most RBS classes have been taught live, where the quality of instruction varies across instructors and classes, and the stability of the implementation of an RBS curriculum is variable. Web-based RBS training may carry important advantages by stabilizing instruction and implementation factors. Randomized trial results of a web-based onsite RBS training program (WayToServe® [WTS]) are reported here. It was hypothesized that servers trained by WTS would refuse alcohol service at significantly higher rates compared with Usual and Customary (UC) live training.

On-site alcohol-serving establishments in New Mexico communities were randomized to receive WTS training (n = 154) or the Usual and Customary live RBS training (n = 155). Premises were assessed at baseline, immediate post-training, 6-months posttraining, and 1-year post-training intervals. Pseudo-intoxicated patron protocols were used to assess premise alcohol service during the early to mid-evening hours of 6:30 P.M. to 8:30 P.M., with the percentage of alcohol service refusals to apparently intoxicated pseudo-patrons as the primary outcome variable.

Results indicate significantly higher refusal rates for WTS than for UC premises at the immediate (WTS = 68% vs. UC = 49%) and the 1-year post-training assessment points (WTS = 68% vs. UC = 58%) but not at the 6-month post-training assessment (WTS = 69% vs. UC = 64%). Differences in refusal rates based on pseudo-patron age were observed where younger pseudo-patrons were consistently refused more often than older pseudo-patrons.

Effective RBS training can be delivered online, making it a potentially cost-effective way of reaching large alcohol server populations.

Alcohol use and misuse is the third leading cause of preventable death in the United States (Mokdad et al., 2004). In 2015, 10,265 people were killed in alcohol-impaired driving crashes (driver’s blood alcohol concentration [BAC] > .08 g/dl; National Highway Traffic Safety Administration, 2016). Drunk driving fatalities (with a BAC > .08) accounted for 31% of all traffic deaths in 2013. Half of all drivers arrested for driving while intoxicated (DWI) drank last in alcohol-serving establishments (e.g., bars, restaurants, and lounges; Chang et al., 1996; Gruenewald et al., 1999; Stockwell et al., 1993; Ventura County Behavioral Health, 2007), and, in one study, more than 10% of binge drinkers drove within 2 hours of their most recent binge drinking occasion (drinking in licensed establishments accounted for 54% of those driving episodes; Naimi et al., 2009).

Alcohol servers may help prevent alcohol-impaired crashes and other harms by discouraging overconsumption among patrons. Although almost every U.S. state prohibits selling or serving alcohol to an intoxicated patron, these laws are not widely enforced (Alcohol Epidemiology Program, 2000; McKnight & Streff, 1994), with alcohol service refusal rates ranging from 11% to 42% (Freisthler et al., 2003; Toomey et al., 1999, 2004, 2005, 2016). These differences in refusal rates may be attributable to a variety of factors, including local law enforcement efforts and the mandatory nature of Responsible Beverage Service (RBS) training.

RBS interventions in the United States typically focus on server training to improve serving behavior. They typically cover (a) state alcohol laws pertaining to a server, (b) procedures for checking patron identification to avoid service to underage individuals, (c) methods for recognizing intoxicated patrons, (d) techniques for preventing over-service (serving too much alcohol on a given occasion), and (e) information on alcohol as a consumable substance.

Evaluations of RBS training have found inconsistent results. Ker and Chinnock (2006), in a Cochrane metaanalytic review, concluded that (a) the effectiveness evidence for server training interventions is mixed; (b) the evidence is complicated by nonrandomized designs, lack of clear outcomes, and incomplete reporting of effect sizes; and (c) consistent implementation of server training is a complicating factor, as the quality and fidelity of RBS training in live classes can vary. Recently Toomey and colleagues (2017) reported a lack of effectiveness for their RBS training program—raising questions about the utility of further research on RBS training (Buvik & Rossow, 2017).

There are two arguments for continuing this research. First, positive evaluations of RBS programs have been reported. For example, the Stockholm Prevents Alcohol and Drug Problems (STAD) program that combined community mobilization, RBS training, and enforcement had substantial effects, increasing alcohol service refusal rates from 5% at baseline to 70% 5 years later (Wallin et al., 2005). It also increased refusal of entrance to apparently drug-impaired patrons (Gripenberg et al., 2007) for up to 5 years (Abdon et al., 2011; although these effects failed to replicate, Skardhamar et al., 2016; Trolldal et al., 2013).

A second argument for further RBS research is that online technologies hold promise for improving training quality and fidelity (and reducing cost), potentially making RBS effects more predictable and cost-effective. Web-based alcohol server training can provide advantages when delivered wide-scale across large populations. It can provide consistent and high-fidelity training implementation by avoiding variation across independent trainers, and its multi-media features can present engaging learning activities and immersive simulations. A number of web-based RBS training courses are commercially available (e.g., learn2serve.com, tipsalcohol.com). However, to date, no systematic test of their effectiveness has been published, comparing it to no training or face-to-face training. Reported here are findings from a randomized trial comparing web-based RBS training for on-premise servers (liquor by the drink)—WayToServe® (WTS)—to usual live training, measured by refusal to serve apparently intoxicated pseudo-patrons.

Furthermore, social factors (e.g., gender and age of both server and patron) were examined for impact on alcohol service refusal. Alcohol service situations may resemble firstimpression situations where social factors affect attributions and decision making (cf. Berger & Calabrese, 1975; Cook, 1979; Kleinke, 1975). Our prior work (Saltz et al., 2001) found that server and pseudo-patron gender affects alcohol sales refusal rates to apparently intoxicated pseudo-patrons. Male servers were significantly more likely to refuse alcohol service to apparently intoxicated female pseudo-patrons than to male pseudo-patrons, and female servers were significantly more likely to refuse alcohol service to apparently intoxicated male pseudo-patrons than to female pseudo-patrons. The age of either the server or the patron has not often been studied specifically in RBS research, although it has been included as a covariate in analyses of RBS laws (Chinman et al., 2013). It has been speculated that, compared with younger servers, older servers (i.e., bartenders) may have more confidence in refusing alcohol sales and supporting RBS practices by coworkers (Fell et al., 2016), although one study (Wallin & Andréasson, 2004) that did examine the age of the server found no difference due to server age when providing service to a young-appearing confederate. Unique to the current investigation, the full range of the age of both the server and the patron as well as the gender of the server and the patron are examined as potential determinants of alcohol service refusal.

Hypothesis and research question

Hypothesis. Alcohol servers in premises trained by WayToServe® (WTS) will more likely refuse service to apparently intoxicated patrons than alcohol servers in premises trained in the usual and customary live RBS training methods (UC).

Research question. Do social factors (i.e., gender and age of servers and pseudo-patrons) significantly affect alcohol service refusal rates?

Before study activities, approval was obtained from the University of New Mexico’s institutional review board. The New Mexico Alcohol and Gaming Division (AGD) provided a list of licensed alcohol-serving establishments. From this list, 309 on-site alcohol-serving establishments were randomly selected from the 566 on-site premises in four New Mexico communities, comprising four types of premises: (a) stand-alone bars, (b) bars that also sold package alcohol, (c) restaurants with a physical bar, and (d) restaurants that served alcohol but did not have a bar. Table 1 shows that the type of premise was balanced between the study arms.

Premise sample by randomized group

When refusing alcohol service to a patron the server should do so RBS

Type of establishmentRandomized group
ControlInterventionTotal
Barn = 42n = 40n = 82
13.7%13.0%26.7%
Bar/restaurantn = 98n = 94n = 192
31.9%30.6%62.5%
Bar/package storen = 0n = 3n = 3
1%1%
Restaurantn = 15n = 15n = 30
4.9%4.9%9.8%
Totaln = 155n = 152n = 307
50.5%49.5%100%

The WTS program conformed to content requirements of the New Mexico Alcohol Server Training curriculum (NewMAST; Saltz & Stanghetta, 1997). WTS was iteratively developed via focus groups with alcohol servers and premise managers in New Mexico that determined proper terminology, idea groupings, and sensitive topics. Once the WTS developed, additional focus groups with alcohol servers and RBS policy makers throughout New Mexico helped to refine its design and structure. WTS was then beta tested in four communities in New Mexico, with five servers per community; all participants found the site to be stable, easy to use, and understandable.

The final version of the website consisted of six multimedia, interactive content modules that included information on skill building, shown in Table 2. WTS emphasized user involvement and interactivity. Each module used video hosts who introduced the module and each of the learning activities: (a) major interactive activities that utilized user involvement to convey information and present and practice RBS skills and (b) “infogadgets” that provided mainly text-based RBS information in a series of tabs anchored by graphics. WTS’s navigation structure was linear, with users required to proceed through modules in order from one to six. To conform to state-mandated testing and certification procedures, each module had an eight-item multiple-choice quiz that users had to pass with at least 80% accuracy to continue and receive an RBS certification from the State of New Mexico. Users who failed a quiz were required to repeat the module. The average completion time for the entire WTS training was 3.71 hours, similar in duration to most live trainings in New Mexico, which are required by state regulation to be at least 4 hours in length.

WayToServe modules and content

When refusing alcohol service to a patron the server should do so RBS

1. Alcohol lawsWhat one can do legally as a server, what one must do as a server, and the penalties one could face for failure to comply with the alcohol laws.
2. Preventing service to minorsIdentifying underage patrons, including the finer points of checking an identification (ID) and information on strategies used by minors in attempting to obtain alcohol.
3. Alcohol and pregnancyInformation about Fetal Alcohol Spectrum Disorder (FASD), including frequently asked questions about FASD.
4. Facts about alcoholThe effects of alcohol and how to recognize a patron who has consumed too much alcohol and information regarding the definition of a Standard Drink and blood alcohol concentration (BAC) levels.
5. The intoxicated patronInstruction and skill-building activity on the active role the server plays in refusing further alcohol service to intoxicated patrons and preventing patrons from driving under the influence of alcohol.
6. Responsible alcohol salesOpportunity to learn and practice the SERVE system for practices sizing up patrons and monitoring their drink consumption to prevent intoxication and to review what has been presented in the previous five sections.

At the time of the trial, servers were required to receive the state-mandated NewMAST in-person training every 5 years. New servers needed to complete training within 30 days of employment. These live trainings, provided by several trainers, conformed to curriculum standards set by the New Mexico AGD and included content on New Mexico alcohol laws, preventing service to minors, alcohol as a substance, alcohol risk during pregnancy, and managing intoxicated patrons. This constituted the UC training condition.

The design was a 2 (level of training: WTS vs. UC training) × 4 (time of assessment: baseline vs. immediate posttraining vs. 6-month post-training vs. 1-year post-training) mixed factorial design. Premises were visited by pseudopatron (PP) assessment teams who conducted the baseline assessment of alcohol purchase attempts before the premises were randomized to the WTS (n = 154) or the UC (n = 155) training condition.

Research staff contacted managers at the premises assigned to the WTS training, secured participation, and registered the servers on the online WTS system. Premises were offered a free project laptop as an incentive to participate. Of the 154 premises contacted for participation in the WTS training, all 154 agreed to participate, with 477 servers registered in the online WTS system in these premises. Two hundred ninety-five servers in the WTS premises completed the WTS training and received a new 5-year New Mexico training certificate. The premises assigned to the UC group were not contacted and remained blind to the study. Attrition of premises due to business closure was 35 premises (11.3%), with more lost in the WTS (n = 24 premises, 15.5%) than in the UC condition (n = 11 premises, 7%).

Immediate post-training PP assessments were conducted approximately 45 days after the intervention sites were given WTS access, followed by the 6-month and the 1-year posttraining assessments. The primary outcome was the rate of alcohol service refusal—how often apparently intoxicated PPs were refused service across establishments.

The PP protocol presented servers with patrons who displayed obvious signs of intoxication. Teams consisted of PP buyers who attempted to purchase alcohol while acting obviously intoxicated at the alcohol-serving establishments and an observer of the purchase attempt who acted as a bystander. Trained observers entered the business ahead of the buyers and rated the fidelity of the performance. The use of PP assessment is a well-established method to evaluate the effectiveness of various alcohol server training models (Gliksman et al., 1993; McKnight, 1989, 1991, 1993; Russ & Geller, 1987; Saltz, 1989).

Pseudo-patrons. Research associates age 21 years or older were trained to behave in accordance with a script comprising such behavior as slurred speech, fumbling with money, lack of eye focus, and difficulty with walking and balance. Raters of videotaped PPs, blind to the script, were 100% accurate in distinguishing the scripted behavior from “sober” presentations. Notably, the script had no threatening or disruptive behavior.

PPs (n = 23; 11 men, 12 women) were recruited through local advertisements and attended five training sessions that explained the project, data collection, and techniques for identifying an obviously intoxicated person. The average age of the PPs was 39.73 years, (SD = 12.74; range = 21–62 years), and 62% were Anglo, 26% Hispanic, and 12% African American. All trained PPs took part in the assessments.

Assessment protocol. The PP teams conducted two visits to each premise between 6:30 P.M. and 8:30 P.M. during each assessment period, separated by approximately 1 week. PP teams did not visit or collect data on the same establishment twice. All premises were blind to the use of the PP protocol, and PP teams were blind to the experimental condition. The choice of the time frame for the assessments (6:30 P.M.–8:30 P.M.) was driven by a number of practical considerations. The relatively early time ensured that restaurants with alcohol licenses were open. Later times would likely correspond with a greater likelihood of intoxication among “real” patrons, and, in some cases, later times might be busier (depending on the specific business). These conditions could affect serving behavior in unknown ways, but the standard time window ensured comparability across types of businesses and time periods.

At the beginning of each PP visit, the observer entered an establishment several minutes before the arrival of the buyer and assessed the premise’s environment (e.g., location of the bar and type of establishment). The observer moved to an area closest to the bar where she or he could observe the buyer upon her or his arrival. At no point during the PP assessment would the buyer or observer acknowledge each other. The buyer arrived acting intoxicated, located a serving area within the observer’s clear view, and engaged the server in conversation before ordering a bottle of Budweiser. If service was refused, the buyer engaged in jovial token resistance but then quickly accepted the server’s decision and exited the premise. If the alcohol was served, the buyer paid for it, excused himself or herself to go to the restroom, and then exited the premise. The observer remained in the premise for up to approximately 5 minutes to observe any comments from the server or other patrons in regard to the buyer, before exiting. Once outside and away from the premise, the buyer and the observer met and completed their data collection forms. In the event of inquiry or confrontation by a premise employee or bystander, the buyer and the observer had a letter explaining that they were part of a research study, although they never used it.

Buyer and observer forms. The buyer and observer recorded both the results of the PP purchase attempt and information on the server (i.e., apparent age, gender, and race/ethnicity) and the premise (i.e., location of alcohol, cleanliness and lighting, placement of required signage, and number of servers present in the premise). Attributed demographics of servers at baseline were 57.9% female, 1.3% African American, 63.8% Anglo, 3.7% Asian, 31.9% Hispanic, 0.7% Native American, with an average age of 31.7 years (SD = 8.52).

The proportion of apparently intoxicated buyers who were refused alcohol service was the primary outcome variable in the analysis. Data were analyzed using SPSS software, Version 23 (IBM Corp, Armonk, NY). Data were analyzed via a mixed model repeated measures analysis of variance (ANOVA; treatment was fixed at two levels [WTS training vs. UC training] and PP assessment was a four-level repeated factor), with planned directional t tests. The interaction between the treatment and the assessment point tested Hypothesis 1. The research question was probed by including the gender and age of the buyers as additional fixed factors in the design and examining three-way interactions among the treatment group, the assessment period, and the social factors. Alpha level was set at p = .05 for all tests. Ancillary analyses were performed that examined the impact of the type of premise.

The 2 (level of training) × 4 (time of assessment) repeated-measures ANOVA on the alcohol service refusal rate found significant main effects for training, F(1, 264) = 5.55, p = .019, and the time of assessment, F(3, 792) = 34.07, p = .0001, and a significant ordinal training by time interaction, F(3, 792) = 2.88, p = .035, partial η2 = .11. Planned t tests (one-tailed critical values were used to test the directional H1, cf. Kirk, 1968) at each assessment period were used to probe the interaction. No significant differences for the refusal rate between WTS and UC training at baseline (WTS refusal rate mean = 33%; UC refusal rate mean = 34%, t(306) = 0.159, N.S.) were found, but at immediate post-training, t(288) = 3.265, p < .0006, Cohen’s d = 0.37, a significant improvement in the WTS (refusal rate mean = 68%), compared with UC training (refusal rate mean = 49%), occurred. At 6-months follow-up, no differences in refusal rates were detected between the WTS (mean refusal rate = 69%) and the UC training (mean refusal rate = 64%, t(282) = 0.794, N.S.). However, at 1-year post-assessment, a significant difference in refusal rates between the WTS (mean refusal rate = 68%) and the UC training (mean refusal rate = 58%), t(272) = 1.739, p = .04, Cohen’s d = 0.21, was detected. Figure 1 displays these means and confidence intervals.

A 2 (level of training) × 4 (assessment point) × 2 (buyer gender) × 3 (buyer age – young adult vs. middle-age vs. mature) ANOVA was conducted. A categorical buyer age variable (young adult = ages 21–25 years, n = 3; middle age = 26–50 years, n = 14; mature age = 51 years and above, n = 5) was included in this analysis as a between-subjects factor. Of the total number of PP visits conducted in the study, 12.3% were conducted by young adult buyers, 53.3% by middle-age buyers, and 34.3% by mature-age buyers. A significant main effect for age category emerged, F(2, 1116) = 22.57, p = .0001, partial η2 = .039.Young buyers garnered more refusals (83.5%) than middle-age (52.4%) or mature (38.2%) buyers. There was no evidence that this social factor moderated the WTS training effect (training × assessment × age, F(4, 1116) = 1.88, p > .05). The gender-related PP main effect and interactions were nonsignificant. The gender and age of server main effects and interactions were also nonsignificant. Finally, analyses examining the impact of the type of premise (bar, bar/restaurant, bar/package store, restaurant) were conducted, with no statistically significant effects found.

WTS RBS training improved refusal rates of alcohol service to apparently intoxicated patrons immediately after training and at 1 year. Given the lack of randomized trials on web-based RBS programs, this is an important finding, which suggests that online training may effectively teach RBS practices compared to live trainings ostensibly following the same state-mandated standardized curriculum. This finding stands in contrast to recent claims about the lack of RBS training effectiveness (cf. Buvik & Rossow, 2017; Toomey et al, 2017) and suggests that RBS training can be effective. Moreover, the level of refusal achieved in the WTS condition (close to 70% at all follow-ups) represented a high level of refusal compared with other published outcomes (e.g., Toomey et al., 1999, 2004, 2005, 2016).

The current investigation does not allow us to separate the impact of web-based delivery from the training content, but it does demonstrate that WTS training was effective. Possible reasons include the following: (a) the WTS training controlled for variability in teaching ability or style and purposeful or inadvertent omission of required content inherent in live training; (b) the WTS program consistently provided standard, high-fidelity training to each trainee across time and location; and (c) by using all media formats (i.e., audio, visual, static and motion graphics, and video), WTS created more engaging, enjoyable, and interactive instruction than live trainers. In particular, online simulations in which servers applied RBS skills in realistic on-premise situations and received feedback to improve them may have benefited servers. Live trainers, using similar techniques, might produce similar results. However, the quality and fidelity of live training varies in practice (Ker & Chinnock, 2006) and raises the question as to whether live trainers can consistently achieve such success. Because of its automation, online training delivery is more consistent and has a better chance of producing similar improvements every time.

The impact of WTS appeared to be immediate and was maintained over a full year, whereas the refusal rate within the UC training group fluctuated. The differences at 1 year, largely due to a decline in refusal by control groups, suggested that the improvement produced by UC training at 6 months was partly due to error variance, whereas improvement achieved by WTS was real and sustained. Randomization should have balanced error variance in UC training, as well as other biases. Future research should replicate persistence of online RBS training effectiveness to determine the best period for renewing training (many states require re-training at 3- to 5-year intervals) and should examine online server training in different alcohol service settings. The effect sizes for WTS versus UC group differences were medium (.37 at immediate post assessment) to small (.21 at 1-year post assessment). These differences may reflect important effects when scaled to an entire state or country, but continued investigation should focus on ways to improve the impact and persistence of web-based RBS training.

Alcohol service refusal rates in the UC condition never exceeded those in the WTS condition, but they did increase at the 6-month post assessment. A number of possibilities might explain this finding. There might have been enhanced enforcement of serving laws by the state regulators around the 6-month follow-up, but a review of agency records showed no extra enforcement. Deviations from the PP protocol could have raised refusal rates in the UC condition, but PP assignments did not differ by condition. Repeated assessment of the premises in a short period (i.e., from baseline to immediate post-training to 6 months) might have increased control premises’ vigilance and refusal rates, a phenomenon witnessed in prior research on alcohol and tobacco products (cf. Erickson et al., 2013; Glanz et al., 2007). Finally, live training in the UC condition might have produced the improvements in alcohol service refusal but achieved less permanency than the WTS training over the 1-year period.

Online delivery of RBS training, such as in WTS, may have greater potential for wide-scale RBS training distribution than live training. It can be easily delivered to large populations over far distances for low cost, when and where servers want it. This may be especially advantageous in the hospitality industry with high staff turnover. However, online training may not be effective for some trainees who benefit from having live trainers respond to their personal needs, monitoring whether they have learned the content, and answering questions. Future research needs to replicate the effectiveness of online training, especially since it has become a common method of RBS training in several U.S. states.

Finally, PP age may influence RBS decisions by servers, with older apparently intoxicated patrons more likely to be served than younger ones. Servers may feel that older patrons are more experienced at managing intoxication and avoiding risky behavior than younger patrons. Servers may have also been more intimidated by older than younger patrons, making refusal more difficult. The influence of customer age should be highlighted in RBS training because servers may not always be aware of it. However, only three PPs were in the young adult category performing 12.3% of the assessments; this finding should be considered preliminary and in need of further investigation.

There were a number of limitations in this study. Conducting PP assessments during later hours at night might have either raised or lowered refusal rates and possibly affected statistical power. This limitation was weighed against practical and safety considerations. Restaurants (with and without bars) comprised more than 70% of the study premises, many of which close after dinner hours. Pseudo-patron visits had to be scheduled during peak hours for those businesses but may have been early for bars that remained open later. Of note, few “late-trading” establishments (those that serve later than 12:00 midnight) exist in New Mexico, unlike in larger urban settings elsewhere. Thus, the findings may generalize best to premises that are not “late trading.” However, if baseline refusal rates had been much higher, concerns might have been raised that PPs were visiting during slow periods and would thus be obvious.

Active recruitment of managers and premises may have overestimated the effect of online training in practice. Managers who participated may have communicated their commitment to RBS techniques and to the importance of the online training more than managers in the UC condition. Management commitment has improved servers’ adherence to RBS principles (Ker & Chinnock, 2006).

PP and premise characteristics may limit the findings. PP assessments might have been affected by patron variability, but the PPs were extensively trained, and different ones made visits to the establishments within and across assessment rounds. Repeated PP assessments could have influenced refusal rates, but randomization should have balanced this across conditions. Differential staff turnover could have affected the results, even in the case of a randomized design. There were no data on the reach or quality of training among the UC training control sites. The study was conducted in a single state that required RBS training of all servers. Whether the findings generalize to all of the 36 states that require, incentivize, or encourage RBS training needs confirmation. Further research should examine the conditions where RBS training is effective at preventing alcohol over-service.

The WTS program is only one form of online training and used interactive media available in the Internet environment, operationalized principles of behavior change and adult learning, and focused on skill acquisition. Not all online RBS courses use multimedia technology, and some focus on presenting knowledge and information more than teaching RBS skills. Whether others are as effective as WTS is unknown and deserves study.

The authors express their appreciation to Drs. Ralph Hingson, Michael Hilton, and Gregory Bloss of the National Institute on Alcohol Abuse and Alcoholism for their support and patience during the conduct of this investigation.

The project took 6 years to complete, from systematic development to testing of the WTS in a multi-community randomized trial in New Mexico. After project completion and a review of our preliminary results, we reported the WTS RBS training program to the Science and Technology Center (STC—the University of New Mexico’s technology transfer center) as an invention, per UNM and NIH rules. At the conclusion of the project, the STC advised our group to commercialize the program via the formation of a private company, and, consequently, Dr. Woodall, Dr. Buller, and Dr. Erwin Bettinghaus (an investor not involved with the research) formed Wedge Communications LLC to do so. The three original primary owners of the program’s intellectual property (i.e., University of New Mexico, Klein Buendel, Inc., and Pacific Institute for Research and Evaluation) licensed the program to Wedge Communications, the State of New Mexico’s AGDcertified WTS, and it became commercially available to servers statewide in 2011. During all phases of the research and commercialization of WTS, our team has complied with the advice, rules, and procedures at the University of New Mexico regarding scientific research and commercialization. Starting with the report of the WTS program to the STC, there has been an active conflict of interest management program ongoing on all research regarding WTS reported by our group, overseen by Barbara McCrady, director of the UNM-Center on Alcoholism, Substance Abuse, and Addictions. The conflict of interest management program involves a regular annual review of all WTS-related manuscripts and presentations. At no time have the researchers or the company taken any funding from any alcohol production, distribution, marketing, or sales organization during the research or commercialization of the WTS program.

This research was supported by National Institute onAlcoholAbuse and Alcoholism Grant R01AA014982 (W. Gill Woodall, principal investigator). Randall Starling, Robert F. Saltz and David B. Buller were co-investigators, and Paula Stanghetta was a consultant to this project.

  • Abdon J., Wallin E., Andréasson S. Long-term effects of a community-based intervention: 5-year follow-up of Clubs against Drugs. Addiction. 2011;106:1997–2004. doi:10.1111/j.1360-0443.2011.03573.x. [PubMed] [Google Scholar]
  • Alcohol Epidemiology Program. Minneapolis, MN: University of Minnesota; 2000. Alcohol policies across the United States: Highlights from the 50 states. [Google Scholar]
  • Berger C. R., Calabrese R. J. Some explorations in initial interaction and beyond: Toward a developmental theory of interpersonal communication. Human Communication Research. 1975;1:99–112. doi:10.1111/j.1468-2958.1975.tb00258.x. [Google Scholar]
  • Buvik K., Rossow I. Server training at drinking establishments: A Sisyphean task? A commentary on Toomey et al. (2017) Journal of Studies on Alcohol and Drugs. 2017;78:276–277. doi:10.15288/jsad.2017.78.276. [PubMed] [Google Scholar]
  • Chang I., Lapham S. C., Barton K. J. Drinking environment and sociodemographic factors among DWI offenders. Journal of Studies on Alcohol. 1996;57:659–669. doi:10.15288/jsa.1996.57.659. [PubMed] [Google Scholar]
  • Chinman M., Ebener P., Malone P. S., Chan K., Imm P., Wright P. A. The carrot and the stick: A cross-sectional study of the influences on responsible merchant practices to reduce underage drinking. Journal of Community Psychology. 2013;41:463–470. doi:10.1002/jcop.21550. [PMC free article] [PubMed] [Google Scholar]
  • Cook M. London, England: Methuen; 1979. Perceiving others: The psychology of interpersonal perception. [Google Scholar]
  • Erickson D. J., Smolenski D. J., Toomey T. L., Carlin B. P., Wagenaar A. C. Do alcohol compliance checks decrease underage sales at neighboring establishments? Journal of Studies on Alcohol and Drugs. 2013;74:852–858. doi:10.15288/jsad.2013.74.852. [PMC free article] [PubMed] [Google Scholar]
  • Fell J. C., Scherer M., Thomas S., Voas R. B. Assessing the impact of twenty underage drinking laws. Journal of Studies on Alcohol and Drugs. 2016;77:249–260. doi:10.15288/jsad.2016.77.249. [PMC free article] [PubMed] [Google Scholar]
  • Freisthler B., Gruenewald P. J., Treno A. J., Lee J. Evaluating alcohol access and the alcohol environment in neighborhood areas. Alcoholism: Clinical and Experimental Research. 2003;27:477–484. doi:10.1097/01.ALC.0000057043.04199.B7. [PubMed] [Google Scholar]
  • Glanz K., Jarrette A. D., Wilson E. A., O’Riordan D. L., Jacob Arriola K. R. Reducing minors’ access to tobacco: Eight years’ experience in Hawaii. Preventive Medicine. 2007;44:55–58. doi:10.1016/j. ypmed.2006.08.021. [PubMed] [Google Scholar]
  • Gliksman L., McKenzie D., Single E., Douglas R., Brunet S., Moffatt K. The role of alcohol providers in prevention: An evaluation of a server intervention programme. Addiction. 1993;88:1195–1203. doi:10.1111/j.1360-0443.1993.tb02142.x. [PubMed] [Google Scholar]
  • Graham K., Bernards S., Osgood D. W., Abbey A., Parks M., Flynn A., Wells S. “Blurred lines?” Sexual aggression and barroom culture. Alcoholism: Clinical and Experimental Research. 2014;38:1416–1424. doi:10.1111/acer.12356. [PMC free article] [PubMed] [Google Scholar]
  • Gripenberg J., Wallin E., Andréasson S. Effects of a community-based drug use prevention program targeting licensed premises. Substance Use & Misuse. 2007;42:1883–1898. doi:10.1080/10826080701532916. [PubMed] [Google Scholar]
  • Gruenewald P. J., Stockwell T., Beel A., Dyskin E. V. Beverage sales and drinking and driving: The role of on-premise drinking places. Journal of Studies on Alcohol. 1999;60:47–53. doi:10.15288/jsa.1999.60.47. [PubMed] [Google Scholar]
  • Ker K., Chinnock P.2006Interventions in the alcohol server setting for preventing injuries Cochrane Database of Systematic Reviews ,2Article No. CD005244 [PubMed] [Google Scholar]
  • Kirk R. E. Belmont, CA: Brooks/Cole; 1968. Experimental design: Procedures for the behavioral sciences. [Google Scholar]
  • Kleinke C. L. Englewood Cliffs, NJ: Prentice-Hall; 1975. First impressions: The psychology of encountering others. [Google Scholar]
  • McKnight A. J. Washington, DC: National Highway Traffic Safety Administration; 1989. Development and field test of a responsible alcohol service program. Volume 3, Final results prevention program (DOT HS 807449) [Google Scholar]
  • McKnight A. J. Factors influencing the effectiveness of serverintervention education. Journal of Studies on Alcohol. 1991;52:389–397. doi:10.15288/jsa.1991.52.389. [PubMed] [Google Scholar]
  • McKnight A. J. Server intervention. Alcohol Health and Research World. 1993;17:76–83. [Google Scholar]
  • McKnight A. J., Streff F. M. The effect of enforcement upon service of alcohol to intoxicated patrons of bars and restaurants. Accident Analysis and Prevention. 1994;26:79–88. doi:10.1016/0001-4575(94)90070-1. [PubMed] [Google Scholar]
  • Mokdad A. H., Marks J. S., Stroup D. F., Gerberding J. L. Actual causes of death in the United States, 2000. JAMA. 2004;291:1238–1245. doi:10.1001/jama.291.10.1238. [PubMed] [Google Scholar]
  • Naimi T. S., Nelson D. E., Brewer R. D. Driving after binge drinking. American Journal of Preventive Medicine. 2009;37:314–320. doi:10.1016/j.amepre.2009.06.013. [PubMed] [Google Scholar]
  • National Highway Traffic Safety Administration. 2015 Motor Vehicle Crashes: Overview; Washington DC: 2016. Traffic Safety Facts. Author. Retrieved from https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/812318. [Google Scholar]
  • Russ N. W., Geller E. S. Training bar personnel to prevent drunken driving: A field evaluation. American Journal of Public Health. 1987;77:952–954. doi:10.2105/AJPH.77.8.952. [PMC free article] [PubMed] [Google Scholar]
  • Saltz R. F. Research needs and opportunities in server intervention programs. Health Education Quarterly. 1989;16:429–438. doi:10.1177/109019818901600310. [PubMed] [Google Scholar]
  • Saltz R. F., Stanghetta P. A community-wide Responsible Beverage Service program in three communities: Early findings. Addiction. 1997;92(Supplement 2):S237–S249. doi:10.1111/j.1360-0443.1997.tb02994.x. [PubMed] [Google Scholar]
  • Saltz R. F., Woodall W. G., Wheeler D. R. Presented at the 2001 Scientific Meeting of The Research Society on Alcoholism; Montreal, Canada: 2001, June. Does gender matter when serving obviously-intoxicated patrons? [Google Scholar]
  • Skardhamar T., Fekjær S. B., Pedersen W. If it works there, will it work here? The effect of a multi-component responsible beverage service (RBS) programme on violence in Oslo. Drug and Alcohol Dependence. 2016;169:128–133. doi:10.1016/j.drugalcdep.2016.10.019. [PubMed] [Google Scholar]
  • Stockwell T., Lang E., Rydon P. High risk drinking settings: The association of serving and promotional practices with harmful drinking. Addiction. 1993;88:1519–1526. doi:10.1111/j.1360-0443.1993.tb03137.x. [PubMed] [Google Scholar]
  • Toomey T. L., Erickson D. J., Patrek W., Fletcher L. A., Wagenaar A. C. Illegal alcohol sales and use of alcohol control policies at community festivals. Public Health Reports. 2005;120:165–173. doi:10.1177/003335490512000210. [PMC free article] [PubMed] [Google Scholar]
  • Toomey T. L., Lenk K. M., Erickson D. J., Horvath K. J., Ecklund A. M., Nederhoff D. M., Nelson T. F. Effects of a hybrid online and in-person training program designed to reduce alcohol sales to obviously intoxicated patrons. Journal of Studies on Alcohol and Drugs. 2017;78:268–275. doi:10.15288/jsad.2017.78.268. [PMC free article] [PubMed] [Google Scholar]
  • Toomey T. L., Lenk K. M., Nederhoff D. M., Nelson T. F., Ecklund A. M., Horvath K. J., Erickson D. J. Can obviously intoxicated patrons still buy alcohol at on-premise establishments? Alcoholism: Clinical and Experimental Research. 2016;40:616–622. doi:10.1111/acer.12985. [PMC free article] [PubMed] [Google Scholar]
  • Toomey T. L., Wagenaar A. C., Erickson D. J., Fletcher L. A., Patrek W., Lenk K. M. Illegal alcohol sales to obviously intoxicated patrons at licensed establishments. Alcoholism: Clinical and Experimental Research. 2004;28:769–774. doi:10.1097/01.ALC.0000125350.73156.FF. [PubMed] [Google Scholar]
  • Toomey T. L., Wagenaar A. C., Kilian G., Fitch O., Rothstein C., Fletcher L. Alcohol sales to pseudo-intoxicated bar patrons. Public Health Reports. 1999;114:337–342. doi:10.1093/phr/114.4.337. [PMC free article] [PubMed] [Google Scholar]
  • Trolldal B., Brännström L., Paschall M. J., Leifman H. Effects of a multi-component responsible beverage service programme on violent assaults in Sweden. Addiction. 2013;108:89–96. doi:10.1111/j.1360-0443.2012.04004.x. [PubMed] [Google Scholar]
  • Ventura County Behavioral Health. Ventura County Place of Last Drink Survey; Ventura, CA: 2007. Alcohol and Drug Prevention Division. Author. [Google Scholar]
  • Wallin E., Andréasson S. Can I have a beer, please? A study of alcohol service to young adults on licensed premises in Stockholm. Prevention Science. 2004;5:221–229. doi:10.1023/B:PREV.0000045356. 37507.86. [PubMed] [Google Scholar]
  • Wallin E., Gripenberg J., Andréasson S. Overserving at licensed premises in Stockholm: Effects of a community action program. Journal of Studies on Alcohol. 2005;66:806–814. doi:10.15288/jsa.2005.66.806. [PubMed] [Google Scholar]