Smartphone apps for managing alcohol consumption: a literature review

Background Smartphone applications (apps) designed to assist users to reduce hazardous and harmful alcohol consumption show potential as an inexpensive alternative to traditional brief intervention in primary care. The aim of this paper is to provide an overview of the literature on alcohol reduction apps and the availability of evidenced-based apps on top commercial app stores. Methods We reviewed literature through to December 2019 using the databases PubMed, MEDLINE, PsycINFO and Google Scholar and keyword search terms smartphone/mobile/phone AND application/app AND alcohol. Articles were included if the primary intervention was a smartphone app and the study measured participant changes in frequency or volume of alcohol consumption. Results 21 relevant articles were identified that evaluated 19 unique smartphone apps. Of the 19 unique apps, seven were designed for use among youth and 12 in adult populations. The available evidence for the efficacy of alcohol reduction apps among youth is inconclusive, with results from these evaluations not showing a clear benefit in reducing alcohol consumption compared to control groups. The results of apps designed for adult populations appears more promising, but results are still mixed. Of the 19 alcohol reduction apps that have been evaluated only eight of these are currently publicly available in commercial app stores. Of these eight apps, only four were demonstrated in the literature to assist with reducing alcohol consumption. Conclusion The evidence for alcohol reduction apps is promising but inconclusive. Few apps that have been evaluated in the scientific literature are currently available for download in commercial app stores.


Introduction
Alcohol use is a leading risk factor for death and disability, accounting for 5.1% of the global burden of disease [1]. Alcohol use has been linked to over 200 health conditions, including cardiovascular disease, cancers, dementia, road injuries, violence and suicides [1]. Despite the risk of adverse consequences, many people still consume alcohol at levels that increase their risk of harm [1][2][3].
Face-to-face brief interventions delivered in primary care appear to be effective in reducing alcohol consumption in patients drinking at hazardous and harmful levels [4]. A 2018 Cochrane review found that hazardous or harmful drinkers who received brief intervention consumed a mean of 20 grams per week less alcohol (95% CI = 28 to 12) after 1 year than participants who received minimal or no intervention [4]. However, several barriers exist which limit the reach of brief interventions in primary care, including heavy clinician workload, a lack of appropriate training and limited time in consultations to manage competing health priorities [5]. For patients, barriers for not seeking treatment include the stigma associated with substance use disorders, and logistical and financial barriers [6].
The use of digital interventions (e.g. websites and smartphone applications), has the potential to overcome some of the barriers by offering an inexpensive, accessible method to deliver education, support and monitoring through personal devices. There have been several metaanalyses conducted evaluating the effectiveness of digital interventions for reducing alcohol consumption that have consistently found small reductions in the intervention groups compared to controls [7][8][9]. A 2017 Cochrane review reported that participants using a digital intervention drank approximately 23 grams of alcohol per week less than controls who received information only or usual care [10].
In more recent years the focus in digital health has shifted to smartphone apps, with more than 318,000 health apps currently available on top app stores and an estimated 200 new apps being added daily [11]. Smartphone apps have several advantages over standard websites, including the ability to be used without internet access and the ability to provide a more personalised experience for users with options for customization. With near ubiquitous smartphone ownership in countries such as the United States [12] and Australia [13], if effective, these apps have great potential to improve health outcomes.
There are a large number of apps currently available in commercial app stores that are aimed at assisting users to reduce their alcohol intake, however few have been evaluated. The aim of this paper is to provide an overview of the literature on alcohol reduction apps and the availability of evidenced-based apps on top commercial app stores.

Method
We conducted a literature review of English language articles published until 4 December 2019 which evaluated the efficacy or effectiveness of a smartphone app in reducing alcohol consumption. The databases searched were PubMed, MEDLINE, PsycINFO and Google Scholar. Keyword search terms were smartphone/mobile/phone AND application/app AND alcohol. Titles and abstracts of all potentially relevant articles were reviewed for possible inclusion by one author (SC). Articles were included if the primary intervention was a smartphone app (native app not a web-based app) and the study measured participant changes in frequency or volume of alcohol consumption. Reference lists from each article were also examined to identify articles that may have been missed in the initial search.
For each unique app identified in the papers we searched for its availability in the top two app stores, Google Play and iTunes in January-February 2020. We searched the United States (US), Canadian, United Kingdom (UK), Irish, Australian and New Zealand (NZ) iTunes and Google Play stores for all apps, plus the iTunes and Google Play stores of the country where the app was developed, e.g. if the app was developed in Sweden we also searched the Swedish iTunes and Google Play stores. If the app could not be found in any Google Play or iTunes stores the corresponding author of the study was emailed to enquire about the app's availability.

Results
The database searches returned 518 results of which 21 were studies evaluating whether use of a smartphone app reduced participants' frequency or volume of alcohol consumption . These 21 articles evaluated 19 unique smartphone apps. Eight of these apps could be found in the iTunes and/or Google Play stores of at least one of the following countries: US, Canada, UK, Ireland, Australia and NZ [15, 19, 20, 22-24, 26, 30, 34]. For the 11 apps that could not be found, corresponding authors of the papers were emailed and three replied: one confirming the app is no longer available in app stores [14,21], one advising the app had been revised and was available under another name [24], and one advising that the app had been temporarily taken down to be updated [17].

Apps for young people
The available evidence for the efficacy of alcohol reduction apps among youth is inconclusive, with results from these evaluations not showing a clear benefit in reducing alcohol consumption compared to control groups [14][15][16][17][18][19][20][21]. Randomised control trials (RCTs) were conducted for five of the seven apps [14-17, 19, 21], with only two of these finding any significant reductions in alcohol consumption outcomes in intervention compared to control groups [16,21]. Of the other two app studies, one was a quasi-experimental, pre-post test design without a control group and found a statistically significant reduction in the proportion of participants reporting binge drinking after the two-week study period compared to baseline (OR = 0.45, 95% CI = 0.37-0.55) [18]. The other, a cluster non-randomized controlled trial, found no statistically significant difference in past month alcohol use between the intervention and control groups (P = 0.014) [20].

Apps for adult populations
Of the 12 apps designed for use in adult populations that have been evaluated, four are aimed at individuals in the Research methods

Self-Record (Jibun Kiroku) [31]
Full-time workers experiencing mild to moderate psychological distress Study design: Non-randomised pilot study with control group Participants: 557 Japanese full-time workers aged 20-59 years who were experiencing mild to moderate psychological distress (Kessler Psychological Distress Scale score between 5 and 12), who indicated an interest in selfmonitoring smartphone apps. Recruited by a research and marketing company Intervention group: 306 participants who self-selected to the intervention group by agreeing to use to the Self Record app. The Self Record app was designed primarily to improve psychological distress, not to reduce alcohol consumption. However, the authors expected that a reduction in psychological distress would lead to reduction in alcohol consumption so tested this, in addition to measuring changes in psychological distress.
The app focuses on self-monitoring and awareness of negative thoughts, daily activities, and daily mood, and provides psychoeducation Control group: 251 participants who did not agree to use the Self Record app. Given only baseline and follow-up surveys, nothing else Follow-up: 4 weeks Intention-to-treat analysis alcohol outcomes: Participants in the intervention group reported increased typical drinking (η 2 = 0.009) and heavy drinking (η 2 = 0.001) compared to controls over the study period. The effect size was small but significant Per-protocol analysis alcohol outcomes: After excluding participants who discontinued using the app after the first day (65% of intervention group) and those who incorrectly answered a validity check item (65% of all participants), results still showed that the intervention group reported significantly increased typical drinking (η 2 = 0.005), and heavy drinking (η 2 = 0.007) compared to those in the control group Availability: Unable to find app in any iTunes or Google Play stores

Results
Cue exposure therapy (CET) [32] Designed as aftercare for individuals who have recently completed an outpatient alcohol treatment program Study design: Investigator-blinded parallel RCT Participants: 164 participants were recruited at the end of a 3-month primary treatment program for AUD at an outpatient alcohol treatment clinic in Denmark. Aimed to compare aftercare treatment approaches CET therapy group: 54 participants randomised to receive group CET therapy with urgespecific coping skills (USCS), conducted by a therapist for 4 × 2 h sessions over 8 weeks.
CET with USCS aims to teach coping skills to reduce cravings, which individuals the practise when exposed to cues to induce cravings CET app group: 54 participants randomised to receive then CET app, which is based on the same treatment principles as the CET with USCS therapy, transformed into a fully automated mobile phone app. Reported BD days reduced from 25.5 days in the 6-weeks prior to 5.8 days during the 6-week study period (P < 0.001). SD consumed/day also reduced from 6.5 to 1.9 SD/day (P < 0.001) Availability: Unable to find app in any iTunes or Google Play stores

D-ARIANNA Digital-Alcohol Risk Alertness
Notifying Network for Adolescents and Young Adults [18] Young people engaging in binge drinking Study design: Natural, quasi-experimental, prepost test design without a control group Participants: 590 participants aged 18-24 years recruited at pubs/clubs and music events Intervention: All participants were asked "did you binge drink in the past two weeks?" with a yes/no response, then were assisted to download the D-ARIANNA app onto their phones and were observed by facilitators to self-administer the app modules at least once. The app estimates current risk of BD by completion of questionnaires and gives a personalised risk estimate indicating how users' risk factors contribute to the overall score Follow-up: Participants were phoned 14 days later and asked whether they had engaged in BD during this period. 83 of the 590 participants could not be reached If the response from the 83 people lost to follow-up were assumed to be the same at the end of the study period as the start, then use of the app was associated with a statistically significant reduction in the proportion of participants reporting BD over the 2-week study period compared to baseline OR = 0.45 (95% CI, 0.37-0.55) Availability: Unable to find app in any iTunes or Google Play stores.
Drinks meter [19] Young people who are current drinkers

Results
Once Upon a High (VoltEgySzer in Hungarian) [20] Young people aged 14-18 years general population who want to reduce their alcohol intake [22,23,26,30], one at working adults experiencing psychological distress [31], and seven at individuals with an alcohol use disorder (AUD) [24, 25, 27-29, 32, 33]. The results from these evaluations are more promising than the apps designed for youth, but results are still mixed. Seven out of 12 studies found significant reductions in alcohol consumption measures, four found no significant reductions and one found a significant increase in alcohol consumption in the intervention group. However, few of these trials were randomized controlled trials with sufficient sample sizes.
Of the four apps aimed at individuals in the general population, two RCTs have been conducted: one found significantly larger reductions in standard drinks consumed per week among participants in one of the app intervention groups compared to controls who were offered the same app with 'active' components removed (P = 0.03) [23]; one did not find a significant difference in standard drinks consumed per week between participants who were offered access to an app and controls who did not receive an intervention (P = 0.17) [22]. Of the other two trials without control groups, one found a significant reduction in Alcohol Use Disorders Identification Test-C (AUDIT-C) scores (P < 0.001) at 3-months follow-up [30], and the other found a significant reduction in alcohol consumption only among 'engaged app users' [26].
Of the seven apps for use among individuals with AUD there have been three RCTs. One found a significant reduction in 'risky drinking days' among participants using an app compared to controls given only treatment as usual (P = 0.003) [25], while the other two trials found no significant differences in any alcohol consumption outcomes between the intervention and control groups [29,32]. The other four trials in this population had mixed results, with three finding significant reductions in alcohol consumption measures and one finding no significant changes.

Availability of apps in commercial app stores
Of the 19 alcohol reduction apps that have been evaluated  as far as we can ascertain only eight of these are currently publicly available in commercial app stores [15, 19, 20, 22-24, 26, 30, 34]. Of these eight apps, only four were demonstrated in the literature to lead to a reduction in alcohol consumption measures [23,24,26,30,34]. Two of these apps are exclusively available in the United Kingdom (UK) iTunes stores [23,34], one is available in the Australian, New Zealand, United States, UK, Irish and Canadian iTunes stores [24], and the other can be accessed from the aforementioned countries' iTunes and Google Play stores but comes at a cost after a 21-day trial period (except for Australian users) [30].

Discussion
Despite the limited number of apps for managing alcohol consumption that have been evaluated in the literature, and the mixed evidence for their efficacy, there are currently hundreds of alcohol-reduction apps publicly available in the top commercial apps stores iTunes and Google Play. The situation is similar for other health issues, such as increasing physical activity [35]. There are hundreds of apps for increasing physical activity, however few are founded in an evidence-base [35].

Commercial app store content analyses
Health researchers have begun to conduct content analyses of commercial app stores in order to identify the highest quality publicly available apps in these stores that could potentially be recommended to health professionals and consumers (e.g. [36][37][38]). These types of analyses involve conducting systematic searches of commercial apps stores to identify publicly available apps purporting to target the issue of interest. Researchers then download the apps and assess them on various quality measures. While this approach is unable to generate evidence regarding the effectiveness of apps, it is a potentially useful method to identify apps that adhere most closely to current evidence-based guidelines and are of the highest technical quality.
There have been several content analyses that have reviewed publicly available apps for managing alcohol consumption [39][40][41][42][43][44], however most do not identify the apps they found to be the highest quality during their analysis specifically by name meaning a reader could not take advantage of these reviews to locate and use the apps identified as the best. One content analysis that does identify the highest quality apps by name is a 2019 study by Tofighi et al., who reviewed the iTunes and Google Play stores for free or low-cost apps claiming to target alcohol or other substance use and evaluated the apps' functionality, aesthetics, and quality of information of using the validated Mobile App Rating Scale (MARS) [44]. The analysis identified the highest rated apps on the scale and provided descriptions and background information on each app [44]. Content analyses such as this could be useful for health professionals and consumers to narrow down the highest quality apps from the hundreds available, and supplement the lack of evidence in this area.
While evidence for the efficacy of apps to reduce alcohol consumption is still in its infancy, and initial results appear to be mixed, established evidence on digital health interventions more broadly does appear to show a benefit  15:17 in reducing alcohol consumption over those receiving no intervention [10]. However, with the paucity of rigorously evaluated apps clinicians could understandably feel hesitant in recommending them to patients. This issue is common for all health-related smartphone applications, with some healthcare professionals and consumers calling for app regulation or certification to be put in place to ensure quality [45]. In lieu of regulation or greater evidence, several groups have developed checklists that aim to assist in assessing the quality of health apps [46,47]. The UK Royal College of Physicians (RCP) Health Informatics Unit has produced a checklist for clinicians to assess the structure, functions, impact and overall quality of health apps [48]. The RCP checklist is designed to help clinicians to feel more confident about recommending health apps to patients [48].

Conclusion
The evidence for alcohol reduction apps is promising but inconclusive. Few apps that have been evaluated in the scientific literature are currently available for download in commercial app stores. In the absence of more scientifically evaluated apps for reducing alcohol consumption, well-designed content analyses of commercial app stores could potentially assist in identifying the highest quality apps that are publicly available.