Edited by: Joel Msafiri Francis, University of the Witwatersrand, South Africa
Reviewed by: Mikkel Magnus Thørrisen, Oslo Metropolitan University, Norway; Claire Garnett, University College London, United Kingdom
This article was submitted to Digital Public Health, a section of the journal Frontiers in Public Health
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
Internet interventions are a promising avenue for delivering alcohol prevention to employees. The objective of this scoping review was to map all research on alcohol internet interventions aimed at employees regardless of design, to gain an overview of current evidence and identify potential knowledge gaps.
We conducted a literature search in three data bases (PsycInfo, PubMed, and Web of Science). Eligibility criteria were that (1) the study targeted employees age ≥18 years; (2) the intervention was delivered predominantly online; (3) the study focused specifically or in part on alcohol use; and (4) the study was published in English in a peer-reviewed academic journal. Two reviewers independently screened, reviewed, and extracted data.
Twenty studies were included, of which 10 were randomized controlled trials, five were secondary analyses, three were feasibility trials, one was a cohort study and one described the rationale and development of an intervention. No qualitative studies were found. Randomized trials tended to show effects when interventions were compared to waitlists but not when more intensive interventions were compared to less intensive ones. We identified two design-related aspects where studies differed; (1) whether all applicants were included regardless of alcohol use level and (2) whether the intervention was explicitly framed as alcohol-focused or not. Significant recruitment problems were noted in several studies.
Alcohol internet interventions hold promise in delivering alcohol prevention to employees, but heterogeneity in study design and difficulties in recruitment complicate interpretation of findings.
Alcohol remains a major risk factor for burden of disease in most parts of the world (
Despite the appropriateness and potential of alcohol prevention aimed at employees, there is a general lack of research in the area, with systematic reviews identifying few randomized controlled trials (
A major development in mental health research in recent decades is the area of internet interventions, i.e., therapeutic interventions delivered
The aforementioned reviews were restricted to RCTs. However, focusing solely on RCTs may provide limited information on the current evidence in a setting such as the workplace where controlled trials are evidently rare and complicated to conduct (
What study designs were used?
What were the characteristics of the workplace setting where the studies were conducted?
How were employees approached?
What types of interventions were used?
What were the demographics of participants?
What were the main findings?
This study was a scoping review attempting to summarize all published literature on internet alcohol interventions aimed at employees, using established methodology (
As this was a scoping review, we did not focus solely on RCTs and other quantitative evaluations but rather included all published research investigating this topic.
In order to be included in the review, the following eligibility criteria had to be met:
The study targeted employees ≥18 years old.
The intervention was predominantly provided over the internet (including mobile apps).
The study focused specifically or in part on alcohol use.
The publication was published in English in a peer-reviewed academic journal from January 2000 onwards (considering the fact that internet interventions were rarely investigated before this date).
Reviews and protocols were not included. In some cases, interventions were simultaneously aimed at several health-related behaviors (i.e., alcohol, physical activity, smoking etc). We then included only those studies where an alcohol-related outcome was reported.
A systematic literature search was conducted, with relevant articles published between January 2000 and February 2021 retrieved from the electronic databases PsycInfo, PubMed, and Web of Science. We combined search terms for “internet,” “alcohol problems,” and “employee” to search titles, keywords, or abstracts. Our search terms in PsycInfo were: (((work* or corporate or white-collar or professional or labor or labor or industrial or staff or organization or organization or employ*)
Once the literature search had been conducted, all duplicates (studies found in more than one database) were removed. Articles were then independently screened in Excel sheets in three stages (title/abstract/full text) by authors CS and DF. At each stage, articles that clearly failed to meet eligibility criteria were excluded. If during the title and abstract screening stage, either of the two screening authors was unsure of whether or not a study should be excluded, that study was retained to be reconsidered in the next stage of screening. Any uncertainty or disagreement concerning whether a study should be excluded during the full text stage of screening was resolved in discussion with author ZK. After the screening process, authors CS and DF extracted relevant data from each eligible article: study characteristics, intervention characteristics and outcome data. Research was then summarized narratively and inductively with a primary focus on responding to the research questions and identifying research gaps.
The combined initial literature search yielded 6,794 articles. Of these, 1,843 were removed as they were duplicates, leaving 4,951 articles for the title screen after which 611 remained. After the abstract screen, 65 remained, and after the full text screen, a total of 18 articles were included in the review. Further, two studies were identified in a hand search, leaving a total of 20 studies. See
Flow of studies through the selection process.
Extraction data from the included studies.
Araki et al. ( |
RCT with three groups: |
Alcohol | Manufacturing plant | 36 (0%) | 43.8 (7.5) | No name of intervention reported |
Drinks in preceding week Alcohol knowledge Attitude toward change |
The face-to-face intervention was significantly more effective than both the internet intervention (e-mails) and the control group in: |
The reason that the face-to-face intervention was more effective may be because the investigator providing the intervention was able to provide tailored feedback, in contrast to the email intervention which was standardized |
Billings et al. ( |
RCT with two groups: |
Alcohol and stress | Technology company | 309 (71%) | 72% were in the 30–49 age range; no mean (SD) reported | “Stress and mood |
Stage of change Work productivity | Those in the internet intervention: |
Authors speculate that the intervention focus on stress makes the program less stigmatizing. Stigmatized behavioral health problems may thus be successfully addressed by embedding prevention material into the stress management framework |
Boss et al. ( |
RCT with three groups: |
Alcohol | Workplace sectors: |
434 (59%) | 47.5 (9.8) | “ |
Drinks in preceding week Emotional irritation Cognitive irritation | When combined, the two intervention groups, compared to the waitlist group, showed significantly larger reductions in drinking 6 weeks after baseline ( |
There were major challenges in recruitment. Authors state that it might be beneficial to collaborate closer with occupational health practitioners and other existing services in order to recruit successfully |
Brendryen et al. ( |
RCT with two groups: |
Alcohol | Four organizations |
85 (52%) | 43 ( |
“ |
Drinks in preceding week | At 2 and 6 months, there were no significant differences between the groups when using a baseline observation carried forward approach (2 months: |
The recruitment potential was overestimated, and the trial was terminated before reaching the required number of participants as defined by the power analysis. |
Buntrock et al. ( |
Secondary analysis of Boss et al. ( |
Alcohol | See Boss et al. ( |
See Boss et al. ( |
See Boss et al. ( |
See Boss et al. ( |
Costs from a societal perspective Costs from an employer perspective | Cost-effectiveness from a societal perspective: the guided intervention had a probability of 55 and 54% respectively of being the most efficient strategy at a willingness-to-pay (WTP) of €0, compared with the unguided intervention and wait-list. At a WTP of €20,000 per QALY gained, the probability was 78% |
This is the only study that investigates cost-effectiveness of internet alcohol interventions among employees |
Colkesen et a. ( |
Cohort study | Health behaviors related to the development of cardiovascular disease | Seven companies, mainly highly educated (i.e., white-collar workers) | 2,289 (39%) | 63% were in the 30–49 range; no |
Initiation of health behavior change (alcohol intake and other behaviors) | Among those receiving alcohol-related feedback ( |
Attrition was high; 6,790 employees were invited. Of these 34% (2,289) participated. Of these, only 28% (638) responded to the 4-week follow-up | |
Doumas and Hannah ( |
RCT with three groups: |
Alcohol | Five local companies | 124 (73%) | All participants were between 18–24; no mean/SD reported | Drinking quantity Peak consumption Frequency of drinking till intoxication |
When combined, the two intervention groups showed significantly larger reductions compared to the control group in weekend drinking [ |
46% of eligible participants participated in the study, and of these 63% participated in the follow-up | |
Hamamura et al. ( |
Feasibility, non-randomized to two groups: |
Alcohol and stress | Recruitment took place |
557 (41%) | 39 (9.6) | “ |
Daily drinking questionnaire | In the intervention group, there were significant increases in typical drinking ( |
Those interested in using the app received the intervention, those not interested received the control. This might explain the negative findings of those receiving the intervention |
Khadjesari et al. ( |
RCT with two groups: |
Alcohol | A large unnamed organization | 1,330 (25%) | 48 (no SD provided) | Number of drinks in preceding week AUDIT EQ-5D (quality of life) | No significant differences between groups were found on any of the outcomes; number of drinks (95% CI −4.7 to 16.9%, |
3% of the employees in the organization participated. It was not clear whether participants actually read the alcohol feedback, particularly as it was presented alongside feedback on other health behaviors | |
Khadjesari et al. ( |
Feasibility study, with a focus on participation and attrition | Alcohol | Six organizations |
1,254 (65%) | 43 (11) | See Khadjesari et al. ( |
Questionnaire on how the intervention was perceived | An average of 8% were recruited from the participating organizations. 25% were drinking at hazardous levels |
A challenge to delivering internet interventions in the workplace is to attract those most likely to benefit; If completion is made compulsory, this may have implications on the veracity of the reported data |
Kouwenhoven-Pasmooij et al. ( |
RCT with two groups: |
Health behaviors | Three organizations; |
491 (19%) | 50.9 (5.8) | Number of drinks per week on a 7-point scale | No differences found between the groups, but an overall within-group decline in excessive alcohol use was observed | The authors speculate that feedback from the intervention may have acted as a warning that subjects needed to improve rapidly, in particular in the military and the police where good physical fitness is a prerequisite for the job. This may explain the finding that improvements were seen in both intervention groups | |
Matano et al. ( |
Study describing the rationale and development of an intervention | Alcohol and stress | One major Silicone Valley work site | No data on employees using the intervention are presented | No data on employees using the intervention are presented | No data on employees using the intervention are presented | No data on employees using the intervention are presented | This paper argues that an easily accessible internet intervention is needed in the workplace. The intervention in question is completely anonymous and is described as simultaneously providing primary and secondary prevention | |
Matano et al. ( |
RCT with two groups: |
Alcohol and stress | One major Silicone Valley work site | 145 (78%) | 39.9 (11.3) | See Matano et al. ( |
Reported alcohol consumption over the past 3 months (quantity and frequency of type of drink, i.e., wine, beer etc) | Medium-risk drinkers: those receiving the full individualized feedback reported a significant decrease in beer binges ( |
The sample size for the targeted group (medium-risk drinkers) was low. Difficulties in recruitment were reported, and may be due to concerns of confidentiality The authors state that the challenge for alcohol prevention at the workplace is to provide a screening tool that offers differential responses depending on the employee's drinking pattern |
Mauriello et al. ( |
Feasibility study | Alcohol | Seven worksites; a university, a research firm, an accounting firm, a wellness center and several restaurants | 166 (69%) | 38.8 (11.2) | Evaluation questionnaire | The post-intervention evaluation questionnaire revealed: |
This is the first intervention that offers tailored feedback on Trans Theoretical Model constructs for each stage of change. When fully developed the intervention will be the first to offer both normative and ipsative feedback on TTM progression | |
Niessen et al. ( |
Secondary analysis of Colkesen et al. ( |
Health behaviors | Five organizations: |
2,686 (46%) | 43.2 (9.5) | See Colkesen et al. ( |
Alcohol consumption scale 0–4 | Those with higher alcohol consumption at baseline were more likely to participate in the study ( |
There was no selective participation related to education, sex or age |
Pemberton et al. ( |
RCT with three groups:1. Drinker's Check-up 2. Alcohol Savvy 3.Waitlist | Alcohol | Eight workplaces—two each from the four military branches |
3,070 (17%) | 69% were in the 21–34 range; no |
1. Drinker's Check-up |
Seven different alcohol measures regarding consumption in the last month: average number of days that alcohol was used, Average number of drinks consumed per drinking occasion and Number of days perceived drunk | At 1 month, Drinker's Check-up had several significant effects on alcohol-related outcomes as compared to waitlist (average days per drinking occasion |
There was considerable attrition (1 month = 66%, 6 month = 77%), which raises the question of response bias. However, the researchers attempted to control for this with sensitivity analyses |
Solenhill et al. ( |
RCT with three groups: |
Health behaviors | Companies in the transport industry | 3,876 (33%) | 43 (10.7) | Number of occasions of alcohol consumption in the past year | There were no significant differences between the groups | The intervention did not have a positive health effect, but an increased short-term motivation to change health behaviors related to diet and physical activity was observed (although this was not the case with alcohol) | |
Westrup et al. ( |
Secondary analysis of Matano et al. ( |
Alcohol and stress | Unidentified worksite | 187 (77%) | 40.9 (11.5) | See Matano et al. ( |
Employee reactions to the website with a follow-up survey | The majority of participants (78%) only used the intervention on 1 day |
Three-quarters of participants were female, which raises the question of whether internet interventions for problem drinking appeal more to women. Participants were mainly highly educated |
Williams et al. ( |
Secondary analysis of Pemberton et al. ( |
Alcohol | Eight workplaces—two each from the four military branches |
3.070 (gender distribution not presented) | Not presented | 1. Drinker's Check-up |
Average number of days |
A number of descriptive norms among those receiving Drinker's Check-up mediated frequency and quantity of drinking, which was not the case with the Alcohol Savvy program | There was considerable attrition |
Williams ( |
Secondary analysis of Pemberton et al. ( |
Alcohol | Eight workplaces—two each from the four military branches |
2,384 | 28.2 (no SD reported) | Drinker's Check-up |
See Williams et al. ( |
Mediation by norms about the number of drinks peers consumed did not vary significantly by gender |
The smaller number of females in the original study may account for the overall lack of effectiveness of norms about drinking frequency as a mechanism of behavior change since its impact was largely confined to women |
The 20 articles reflected 16 unique studies and 13 projects; (see
Of the 20 studies identified, 10 were randomized controlled trials (
Of the 20 included articles, 11 were conducted in several workplaces simultaneously (
There was large variation both between and within studies in terms of sectors. In the six articles reporting on studies conducted in a single workplace, information about sector was sparse; only two revealed the sector of the workplace where the study was being conducted (
The most common way of informing employees about the study was through employee emails or web portals belonging to the workplace(s) where the study was being conducted (
The content of the interventions in the studies fell into two main categories; personalized normative feedback (PNF) and cognitive behavior therapy (CBT). PNF is a form of brief intervention aiming to inform individuals about their consumption patterns in comparison to those in the general population (
Most interventions included PNF. Some of these interventions offered PNF for alcohol only (
Of the interventions based on CBT, two had an explicit focus on alcohol consumption (
Four studies included some form of human contact in one of the groups. In one study, the intervention was delivered together with optional guidance from a therapist over the treatment platform (
There was variation between the studies, with some including a large majority of females (
A host of different alcohol outcomes were used in the studies of which the most common one was number of standard drinks consumed or some similar calendar-based way of assessing recent consumption (
In terms of intervention effects in the RCTs, seven studies used wait-lists as control groups to assess effects. In four of these, the internet interventions were found more effective than the waitlist in at least one alcohol-related outcome (
Some studies also investigated mental health outcomes. Billings et al. (
Only two studies reported work-related outcomes: Billings et al. (
One study described the rationale and development of an intervention, building the case for developing an alcohol internet intervention that can simultaneously offer primary and secondary prevention (
When synthesizing our results, we identified two main differences in study design in the RCTs, that may bear consequences for interpretation of results. Below, we formulate these differences as questions (see
Study design aspects of the 10 RCTs.
Araki et al. ( |
No (cut-off: GTP) | Yes |
Billings et al. ( |
Yes | No |
Boss et al. ( |
No (cut-off: drinks in preceding week) | Yes |
Brendryen et al. ( |
No (cut-off: AUDIT score) | Yes |
Doumas and Hannah ( |
Yes | Yes |
Hamamura et al. ( |
Yes | No |
Khadjesari et al. ( |
No (cut-off: AUDIT score) | No |
Kouwenhoven-Paasmoiij et. ( |
Yes | No alcohol focus |
Matano et al. ( |
Yes | No |
Pemberton et al. ( |
Yes | Yes |
Solenhill et al. ( |
Yes | No alcohol focus |
In six RCTs and one non-randomized trial, all employees were welcome to participate in the study regardless of their level of alcohol use (
In some studies, participants were explicitly informed about the alcohol focus of the intervention, while in other studies the focus on alcohol was intentionally masked. This was done, for example, by framing the intervention as how to deal with stress (
Recruitment problems were reported in several studies (
In this scoping review, we identified 20 studies about alcohol internet interventions in the workplace. Half of the studies were randomized controlled trials, five were secondary analyses, three were feasibility studies, one was a cohort study and one described the rationale and development of an intervention. No qualitative studies were identified. The interventions consisted of personalized normative feedback and/or cognitive behavior therapy. The gender distribution among participants was similar in most studies and the average age was similar to other alcohol internet intervention studies (
The field of alcohol internet interventions is small and the trials conducted to date have evidently encountered many challenges. Improved methodological stringency paired with an awareness of the unique challenges this context provides is needed. Below, we discuss some targeted efforts that the field would benefit from at this stage.
First, the education level of those participating was not reported in most studies. This should be studied in greater detail, especially considering the fact that some research suggests that education level is negatively associated with outcomes in alcohol internet interventions (
Detailed information about demographic characteristics such as education level and profession, should be collected, as this might aid in understanding which participants benefit most from these interventions.
In order to facilitate comparisons of effects across studies, researchers should include the most common alcohol-related outcome measures in the field.
Work-related outcomes should be included to a greater extent.
Qualitative studies would likely help in understanding how studies could be designed to decrease problems in recruitment.
Researchers should be conscious of how design decisions regarding inclusion criteria and framing of the intervention may influence what conclusions can be drawn.
Future studies should include active controls rather than wait list controls.
Studies conducted in low- and middle-income countries are sorely needed.
This review made use of a broad search strategy and we conducted searches in three widely used databases. Although these databases are large, it is possible that some studies were not identified in this process. Also, we did not search for gray literature. Scoping reviews generally are limited in what can be generalized, in comparison to systematic reviews.
Alcohol internet interventions hold promise in delivering alcohol prevention in the workplace. However, heterogeneity in study design and difficulties in recruitment complicate interpretation of findings. The field would benefit from a consensus statement on issues related to study design and outcomes.
The original contributions presented in the study are included in the article/
CS: conceptualization, data curation, formal analysis, funding acquisition, investigation, methodology, project administration, writing—original draft, and writing—review and editing. DF: conceptualization, data curation, formal analysis, funding acquisition, investigation, methodology, project administration, and writing—review and editing. AB: funding acquisition, investigation, and writing—review and editing-equal. ZK: conceptualization, formal analysis, investigation, methodology, and writing—review and editing. KS: conceptualization, formal analysis, funding acquisition, investigation, methodology, project administration, writing—original draft, and writing—review and editing. All authors contributed to the article and approved the submitted version.
This work was funded by AFA Försäkring. The funder was not involved in the study design, collection, analysis, interpretation of data, the writing of this article, or the decision to submit it for publication.
CS has been on the clinical review board of the digital mental health company Alavida. The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
The Supplementary Material for this article can be found online at: