Project CHOICE: Choosing healthy options in coping with emotions, an EMA/EMI plus in-person intervention for alcohol use

Abstract

Background:

Emerging adulthood (18–25 years) represents a risky time for mental health and substance use. Emerging adults are particularly susceptible to problematic patterns of substance use, especially if they experience anxiety and/or depression and use substances as a way to cope with such issues. However, many mental health treatments do not address substance use. We developed an ecological momentary assessment and intervention (EMA/EMI) to specifically target the motive of drinking to cope with anxiety/depression.

Methods:

Project CHOICE was a 6-week intervention that paired in-person normative feedback with daily EMA and, if an individual reported negative affect and intent to drink, an EMI was immediately sent to their phone (a personally-chosen coping skill). We recruited n = 20 (55% female, mean age 21.74, 85% Caucasian and 75% non-Hispanic/Latino) individuals from a psychiatric partial hospitalization program for a 6-week open trial of the CHOICE intervention and re-assessed at the 6-week follow-up point.

Results:

Results indicated that drinking variables and coping motives were highly correlated at baseline. Days of drinking, alcohol-related problems, and coping motives significantly decreased over time following the intervention. Results indicated high levels of feasibility and acceptability.

Conclusions:

This open pilot represents a feasible, acceptable, and promising direction in delivering interventions in the moment when risk is highest, utilizing smartphone capabilities.

Introduction

Alcohol use and mental health issues are common and problematic among emerging adults (EAs). Rates of alcohol use are highest among this age group, with close to 60% reporting current drinking and 10.7% meeting past year alcohol use disorder (AUD) diagnostic criteria, a rate nearly double the U.S. average of all other age groups.¹ Among EAs, rates of anxiety and depression are also high, with over 20% reporting past year mental illness.¹ Notably, because negative affect is a cue for alcohol use,^2,3 individuals with anxiety and depression are significantly more likely to use alcohol to cope and also to abuse alcohol than those who do not experience anxiety and depression.^1,4 Indeed, those who use alcohol specifically for the coping motive (i.e., drinking to cope with negative affect) are significantly more likely to have more alcohol-related problems as compared to individuals who drink for other motives.^5–9 Developing and testing interventions for EA who are at risk for drinking to cope—such as those in psychiatric treatment—may result in a significant public health impact.

Despite the fact that alcohol use exacerbates anxiety and depression^3,10 and drinking to cope is associated with problematic outcomes,⁸ existing treatment for anxiety and depression often do not address alcohol use.¹¹ Additionally, alcohol interventions for EAs largely ignore underlying depression and anxiety which may contribute to use. Meta-analyses indicate that effects of alcohol interventions for EAs are modest,^12,13 perhaps because few interventions have focused on psychiatric comorbidity or on the coping motive. Indeed, the few studies that have focused on specific underlying motives for drinking have shown promising results in reduction of alcohol use and negative consequences.^14–16

Building off of previous research and interventions designed to reduce the coping motive, we previously developed and implemented a brief, personalized feedback intervention (PFI) that focused on reducing problematic alcohol use by specifically targeting use of alcohol to cope among a normative sample of EAs (PFIcope).¹⁷ Participants (n = 170) were randomly assigned to receive either the PFIcope intervention (n = 87) receiving normative feedback on their alcohol use, education and feedback on the coping motive and their use of alcohol to cope with alternate coping strategies, or to a standard feedback condition (n = 83) which received normative feedback on alcohol use only. At the 2-month follow-up, results indicated that individuals who received PFIcope reported significant reductions in drinking to cope. These findings suggest that, with adaptation for needs of a clinical population, the PFIcope study could be an ideal intervention for EAs in psychiatric settings who drink to cope and are at risk for problematic outcomes related to their alcohol use.

Harnessing the advantages of technology-supported approaches, the potential efficacy of the in-person PFIcope intervention for a psychiatric clinical population with comorbid alcohol use may be further optimized. Specifically, mobile and text-based interventions for substance use disorders are acceptable, feasible, may improve access to care, increase engagement and retention,^18–20 and have shown promising results in the reduction of alcohol use.^21,22 Although it is important to help an individual monitor alcohol use, interventions that promote relapse prevention (e.g., use of coping skills) efforts are more effective at reducing use than self-monitoring alone.^23–28 Despite the proven effectiveness of technology-based treatment, it may be difficult to ascertain when an individual is in a high-risk situation for alcohol use and in most need of a prevention intervention. The use of Ecological Momentary Assessment (EMA) has been commonly-used as an electronic method to monitor use and measure factors that lead to use, high-risk drinking situations such as negative affect, which activates the coping motive, and intent to drink.^29–32 Effective interventions can utilize EMA to trigger a real-time intervention (Ecological Momentary Intervention (EMI)³³) delivered when an individual is at greatest need. Thus, bolstering the PFIcope intervention with EMA to identify “drinking-to-cope” high-risk situations in real-time, paired with a just-in-time intervention (EMI) that offers individualized, coping-focused relapse prevention strategies, represents a novel, theory-based treatment.

We developed a 6-week personalized feedback plus EMA/EMI intervention, Project CHOICE (Choosing Healthy Options in Coping with Emotions), for EAs in a partial hospitalization program who drink to cope with negative affect. This study had two aims: to determine the feasibility and acceptability of Project CHOICE (as assessed by EMA completion, participant interview, and attrition rates) and to evaluate the impact on drinking motives and alcohol outcomes. We hypothesized that individuals who engage in Project CHOICE would report reductions in alcohol use and drinking to cope.

Methods and materials

Participants and procedures

Participants (n = 20) were recruited from an EA partial hospitalization program at a large, private psychiatric institution. The program runs from 9:00 am to 3:30 pm, Monday through Friday, for approximately 7 days, serving EA for a variety of psychiatric problems, most commonly depression and anxiety. Many EAs in the program also have co-occurring substance use problems which, although they are assessed at the time of entry into the program, specific substance use programing is not offered. Program components include group therapy with 3–4 topics covered daily, individual counseling, and evaluation by an attending psychiatrist (who may prescribe psychiatric medication).

Recruitment occurred between September 2018 and February 2019. Patients’ medical records were reviewed by study staff in accordance with Institutional Review Board guidelines to assess the inclusion and exclusion criteria. Additionally, participants were screened in-person to verify eligibility. Inclusion criteria included: being between ages of 18–25, past month alcohol use greater or equal to 2–3 times weekly, a score of 2+ average on the coping scales of the motives measure (see below), current anxiety and depression symptomatology (as measured by scores of 6+ Generalized Anxiety Disorder-7³⁴ corresponding to moderate to severe anxiety, and a score of 16+ on the Center for Epidemiology Studies-Depression scale,³⁵ corresponding to high risk for clinical depression), and smartphone ownership. Exclusion criteria included current diagnosis of other moderate/severe substance use disorder (other than alcohol), history of or current psychotic disorder, and current suicidal/homicidal ideation. Out of the 347 individuals admitted to the program, 304 were not approached for the following reasons: current DSM-5 diagnosis of other moderate/severe SUD (n = 26; the vast majority being cannabis use disorder), history of psychotic disorder (n = 22), current SI/HI (n = 18), no show/discharged from the program (n = 11), no alcohol use or AUDIT <3 (n = 208), and other issues such as staff availability, patient age, or current mania (n = 19). Patients appearing to meet eligibility were provided with a brief study description, and screened if interested. Of the 43 individuals screened, n = 3 ruled out for psychiatric issues, n = 14 individuals canceled program participation (i.e., left YAPH before they had a chance to be enrolled) or were not interested in the study, n = 5 did not meet inclusion criteria upon further screening. A total of n = 20 individuals were interested, eligible, and completed a baseline assessment followed by a study orientation session. Participants were compensated for baseline and follow-up ($20 per session), and for completion of EMA prompts at $0.50 per response for a maximum of $90 EMA compensation; thus, participants could receive up to $130 for participation.

Intervention components

This 6-week open pilot was comprised of the following components:

Orientation session

This in-person session with a PhD-level clinical psychologist included: (a) normative feedback on their alcohol use compared to other EAs, (b) education on the coping motive and negative outcomes associated with use, (c) normative feedback regarding the individual’s use of alcohol to cope (based on responses to the motives measure as compared with normative samples), (d) generation of alternate coping strategies to inform EMI text relapse prevention messages (e.g., try taking a walk if you feel like drinking, calling a friend has helped you in the past when you feel down or want to drink), (e) installation of the smartphone app (mEMA, designed by ilumivu) on the individual’s smartphone, and (f) orientation to components and timeline of the intervention, including a discussion of methods of increase EMA adherence/compliance. Components a–c were used in the previous PFIcope study.¹⁷

EMA prompts

Following discharge from YAPH, participants received EMA prompts 4x/daily (sent at random times between the hours of 9 am–12 pm, 12 pm–3 pm, 3 pm–6 pm, and 6 pm–9 pm) through the mEMA app, a HIPAA-compliant web-based platform specifically developed for EMA. We assessed: (A) current mood at that moment (“RIGHT NOW;” stress, sadness, happiness, excitement, and anger, rated on a scale Likert scale of 1–10), alcohol consumption since the previous assessment (yes/no, quantity), drinking intention (yes/non), drinking motives (if drinking was reported, utilizing one item from each subscale of the DMQ-R; enhancement, coping, social, and conformity), the use of coping strategies (14 strategies provided endorsed yes/no; e.g., distraction, active coping), any cannabis use since the previous assessment (yes/no), use of cannabis for substitution of alcohol (yes/no) and motives for cannabis use. A total of 14 questions were included in the EMA. EMA prompts were also available on demand if a participant intended to consume alcohol or reported negative affect and desired the EMI support.

EMI relapse prevention coping skills messages

The mEMA app allows for individualized programing. As such, messaging can be specifically designed for each participant. If the participant endorsed negative affect or intent to drink on any EMA prompt, an individualized, self-chosen coping skills-based message was immediately sent through the EMA app. Participants picked their individualized coping skills messages in the orientation session from a list of coping strategies that are consistent with cognitive behavioral therapy techniques, and were also encouraged to add their own strategies if they were not listed on the sheet (e.g., “Try reaching out to a friend or family member if you feel down or want to drink,” and the 10–12 messages chosen were randomized to be sent in the event of negative mood and intent to drink. The purpose of these messages was to remind the participant of skills and strategies to reduce the likelihood of drinking to cope.

Follow-up assessment

At the end of the intervention (6 weeks), participants completed self-report questionnaires and a semi-structured, in-person qualitative interview at the research unit. The purpose of the interview was to determine which aspects of the intervention were helpful and elicit other feedback in order to shed light on acceptability, usability, and utility of the intervention.

Measures

Feasibility and acceptability

We operationalized feasibility and acceptability via recruitment rates, EMA response rates, follow-up rates, and qualitative interviews. Additionally, at each EMA prompt, participants were asked whether they used a coping skill because of the message they received on the app (yes/no).

Motives

Alcohol use motives were measured using the Modified Drinking Motives Questionnaire-Revised (MDMQ-R),³⁶ a 28-item assessment derived from the Drinking Motives Questionnaire-Revised.⁵ The MDMQ-R measures the following motives: coping with anxiety (“because it helps me when I am feeling nervous”), coping with depression (“to cheer me up when I’m in a bad mood”), conformity (“to fit in with a group I like”), social (“as a way to celebrate”), and enhancement (“to get a high”). Participants are asked to rate on a scale of 1 (almost never/never) to 5 (almost always/ always) how frequently they drink for each reason. Alpha coefficients are as followed for baseline and follow-up: coping with anxiety (α = 0.89 and 0.79), coping with depression (α = 0.97 and 0.94), conformity (α = 0.61 and 0.82), social (α = 0.55 and 0.79), and enhancement (α = 0.84 and 0.73).

Problematic drinking

Problematic drinking was assessed using the Rutgers Alcohol Problem Index (RAPI).³⁷ The RAPI is a 23-item self-report measure of problems associated with use. Participants rate on a scale of 0 (none) to 3 (more than 5 times) how often they experience alcohol-related problems in the previous month. Alpha reliability coefficients ranged from α = 0.87 at baseline to α = 0.77 at follow-up.

Alcohol and cannabis use

The Timeline Followback (TLFB)³⁸ was utilized to measure quantity and frequency of alcohol use. Participants are given anchor points and produce their own anchor points (including holidays, special occasions) to help facilitate memory of alcohol use, and asked to recall alcohol use (drinks per drinking day, DDD). We calculated binge days according to National Institute of Alcohol Abuse and Alcoholism cutoffs (>3 drinks per day for women and no more >4 drinks per day for men). We also derived a drinks/week variable from the data. Additionally, on each drinking day, participants were asked to report (yes/no) whether their alcohol use was motivated by the coping motive. As such, we created a variable that was the sum number of days that an individual endorsed drinking to cope on the TLFB, which allowed for computation of percentage of drinking days that an individual drank to cope. We also assessed days of cannabis use using the TLFB. At baseline, the previous 90 days were assessed while at the 6-week follow-up days since baseline were assessed (42 days).

Data analysis plan

Preliminary analyses evaluated descriptive statistics of variables of interest for missingness and handled such instances accordingly.^39,40 Variables were examined for normality and analyses accounted for skewness or kurtosis utilizing boot-strapping methods or transformations. We first present descriptive statistics and correlations between variables, as well as results from analysis of variance (ANOVA) testing examining differences in baseline variables by gender. Main effects of the intervention were analyzed using paired-sample t-tests utilizing measures at baseline and 6-weeks. The primary outcomes of interest were days of alcohol use and DDD (as assessed by the TLFB), binge drinking, problematic drinking, and motives. Secondary outcomes were changes in motives (specifically drinking to cope). We utilized two measures of drinking to cope (a) the measure of the extent to which individuals utilized drinking to cope at BL and 6 W (using the MDMQR³⁶) and (b) a daily frequency count of using to cope during drinking days endorsed on the TLFB (yes/no, see Abrantes⁴¹) Qualitative interviews were audiotaped and transcribed by a research assistant and through n-Vivo software’s electronic transcription capabilities (QSR International) in order to verify accuracy. The first author identified themes from the transcribed interviews, and the second and fifth author reviewed the themes until consensus was reached.

Results

Descriptive statistics

Participants (n = 20) had a mean age of 21.74 (SD = 2.23) years of age primarily female (55%), Caucasian (85%), and non-Hispanic/Latino (75%). The sample was diverse in terms of occupational status: 40% were unemployed, 40% were employed, and 20% were students. The majority of participants lived with parents (65%). Of the 20 individuals enrolled, 15 completed the follow-up survey and associated interviews while 8 also completed the qualitative interview. Ten participants began a medication protocol while in the program. At follow-up, six participants were still taking baseline psychiatric medications for anxiety or depression with no changes to dosage, 2 participants stopped taking psychiatric medications, and 2 participants had medications adjusted since baseline. No participants were administered medications for the treatment of alcohol use disorder.

On average, participants were drinking 35% (SD = 0.25) of the previous 90 days at baseline with an average 3.5 (SD = 1.67) drinks per drinking day, 8.80 drinks per week (SD = 5.52), and 44% of drinking days (SD = 0.35) were reportedly days on which participants drank to cope (Table 1). Participants reported binge drinking on 21.7% (SD = 0.160) of days at baseline. Participants endorsed a mean score of 13.2 (SD = 10.8) on the RAPI. The most highly-endorsed motives were coping with anxiety (mean = 3.45, SD = 1.14) and enhancement (mean = 3.37, SD = 0.93), followed closely by social motives (mean = 3.30, SD = 0.43) and coping with depression (mean = 2.85, SD = 1.49). Participants endorsed coping to conform at the lowest levels (mean = 1.60, SD = 0.60). Sixteen participants reported cannabis use at baseline, while only 10 reported cannabis use at follow-up. Baseline variables did not significantly differ by gender (all ps > .05). Data from the EMA indicated that participants reported using distraction techniques most frequently, followed by active coping.

Table 1.

Descriptive statistics and change over time of use, problematic drinking, and motives.

	Baseline		6 Weeks		Change over time
	Mean	SD	Mean	SD	t	p
% Days use	0.35	0.25	0.19	0.22	3.27	<.01
% Days coping	0.44	0.35	0.16	0.25	3.09	<.01
Drinks per drinking day	3.56	1.67	2.91	0.22	1.96	.07
Binge drinking (%)	21.70	16.06	8.25	11.09	2.82	.01
Drinks per week	8.80	5.52	2.91	3.73	4.57	<.01
Problematic drinking	13.23	10.83	4.31	4.50	3.02	.01
Social	3.30	0.43	2.72	0.52	3.49	<.01
Coping with anxiety	3.42	1.14	2.35	0.87	2.85	.02
Coping with depression	2.85	1.49	1.97	0.84	2.16	.05
Enhancement	3.37	0.93	2.83	0.86	1.77	.10
Conformity	1.60	0.60	1.19	0.30	2.93	.01

Note: t-tests were performed on the n = 15 individuals who completed follow-up.

Correlations

Table 2 describes correlations at baseline (below diagonal) and at follow-up (above diagonal). Results at baseline indicated that percentage of days that individuals drank to cope was associated with binge drinking (r = 0.66), problematic drinking (r = 0.66), motives of coping with anxiety (r= 0.84), coping with depression (r = 0.87), and enhancement (r = 0.45). Drinks per drinking day were associated with binge drinking (r = 0.56), problematic drinking (r = 0.64), coping with depression (r = 49), and enhancement (r = 52). Problematic drinking was associated with coping motives (anxiety r = 0.63, depression r = 75) and enhancement (r = 0.54). Several motives were highly-correlated (especially coping and enhancement). At follow-up, percent days drinking was associated with percent days of coping (collected data and implemented r = 0.70). DDD were associated with social motive (r = 0.62).

Table 2.

Product moment correlation coefficients.

	1	2	3	4	5	6	7	8	9	10
1. % Days use	_	.070**	0.05	0.70**	−0.31	−0.31	0.26	0.03	0.12	0.03
2. % Days coping	0.05	_	−0.08	0.26	−0.28	−0.51	0.56	0.18	0.13	−0.12
3. Drinks per drinking day	−0.02	0.41	_	0.49	0.21	0.62*	−0.04	−0.10	−0.13	−0.22
4. % Binge drinking	0.53**	0.45*	0.56**	_	−.26	0.28	0.05	−0.15	0.06	−0.24
5. Problematic drinking	−0.12	0.66**	0.64**	0.35	_	−0.21	0.40	0.46	0.37	−0.26
6. Social	0.21	0.25	0.33	0.01	0.18	_	−0.03	0.75**	−0.23	−0.04
7. Coping with anxiety	0.20	0.84**	0.34	0.43	0.63**	0.41	_	0.74**	0.74**	−0.27
8. Coping with depression	−0.07	0.87**	0.49*	0.37	0.75**	0.30	0.87**	_	0.67*	0.22
9. Enhancement	0.13	0.45*	0.52*	0.20	0.54*	0.58**	0.57**	0.58**	_	−0.21
10. Conformity	0.17	0.15	0.08	−0.07	0.11	0.64**	0.40	0.37	0.18	-

^*p < .05;

^**p < .01.

Note: below the diagonal are baseline correlations, above the diagonal are follow-up correlations.

Changes over time

Several variables of interest showed significant change over time (see Table 1). Participants significantly reduced percent days of alcohol use (t = 3.27, p< .01), percent days of binge drinking (t = 2.82, p= .01), percent days of use for the coping motive (t = 3.09, p< .01), problematic drinking (t = 3.02, p= .01), coping with anxiety (t = 2.85, p= .02) and coping with depression (t = 2.16, p= .05), social (t = 3.49, p < .01), and conformity motives (t = 2.93, p= .01).

Feasibility and acceptability

Participants completed 78% of EMA prompts during weeks 1–2, 71% during weeks 1–4, and 62% during weeks 1–6. 75% of individuals completed follow-up assessments. Participants reportedly used coping skills that were specifically suggested by the app as assessed after each EMA prompt over half of the time that they responded to the 4xd/daily prompts (52.2%). 75% of individuals reported that they thought that the app provided an appropriate and helpful number of messages, while 12.5% of individuals thought that messages were too frequent and 12.5% felt as though messages were not frequent enough. However, some participants reported in the qualitative interview that they found it difficult to complete all of the EMA prompts: “What made it difficult for me was the fact that if you didn’t do the check in within that hour you like lost the opportunity to do it whatsoever.” Additionally, 3 participants reported in the qualitative interview that they wish they were offered more coping strategies. All participants (100%) reported that they would continue to monitor their drinking to cope in the qualitative interview and all participants reported that they believed the intervention was helpful. Several themes emerged in the qualitative interviews: identifying and tracking emotions, insights about their drinking habits, the role of emotions in their drinking, the use of coping skills, and other benefits of the app (see Table 3).

Table 3.

Qualitative interview themes.

Identifying and tracking emotions

• I just liked how [the app] checked in all the time with me and help give me coping skills to use. It also helped me notice how I was feeling to how to cope with emotions.

• It literally just made you really think about what was going on and how you actually were feeling across the day. So I guess I liked to really just check how I was doing and I was like “Is it even worth it to feel this way right now?”

• I liked the random check ins when it asked me how I was feeling throughout the day. A problem that I have is I really am not as in tune to how I’m feeling and [the app] really helps me to feel like “oh wow you’re really anxious right now you should probably do something about that.”

Insight into drinking patterns

• [The app] completely keeps you in check about these using substances.

• It made me more conscious of my drinking habits. And it made me realize the amount that I was drinking.

• I definitely think that I had been going down a bit of a slippery slope for a bit and I think I’m doing a lot better now.

• I know I can sometimes use [alcohol] as a bit of a crutch.

The role of emotions in alcohol use

• When I was doing the study I was kind of putting two and two together … grouping [emotions and drinking] together more makes me realize that I would drink more because I was sad and that’s not good.

• [The app] opens your eyes to how quickly mood can change. And that mood can correlate to certain habits in your life.

• I actually have not used alcohol nearly as much as you did prior to even entering the young adult program and starting this study. I think it’s just because I was more attuned to how I was feeling and I actually questioned what is my reasoning behind taking the drink: is it because I’m actually out with friends having a good time or am my trying to cope with negative emotions?

• Oh yes [I will continue to monitor my drinking to cope] because a lot of the times when I’m drinking or doing these like negative behaviors, I’m doing it and I don’t even know why.

The use of coping skills

• Just seeing [a coping idea] on the screen and having it pop up and remind you of a good coping strategy really helped because half the time I can’t do that on my own.

• And so even though some days I was it really even feeling like drinking and also just helped me on days where I was just actually struggling with my emotions in general so I could just be like “Oh I’m feeling really anxious right now” and [the app] would pop up and tell me to go out and exercise.

• I learned that I need to be more in tune with myself and I learned that I have a lot of great coping skills and when I’m when I have negative feelings, I can get myself out of that … And it was great to have this reminder - I have this phone with me 24/7 and it’s such a helpful tool to have this app.

• I think it’s important to be able to check in with myself and say “why am I doing this? Is there a reason behind it? Am I anxious and depressed?” … So it’s this is really kind of just taught me to monitor my feelings and then to employ positive coping skills rather than going for those negative maladaptive coping skills.

Other benefits of the app

• [The app] helps you live in the now like a hundred percent.

• Some of my family are alcoholics. Seeing them, I just always said to myself I never want to get kind of down to that low … I don’t want to go down that path.

• I really would recommend Project CHOICE because it really helped me with my drinking. Before, I didn’t have anyone really telling me to slow down or saying that’s enough.

Discussion

Although drinking to cope appears to be fairly common among emerging adults in psychiatric care, future interventions should also address other drinking motives, even amongst a highly-specific group of individuals. Participants were drinking over one-third of days at baseline, and almost half of those days were endorsed as days that they drank to cope. Additionally, participants were drinking over 3.5 drinks/day, a level much greater than recommended by the National Institute of Alcohol Abuse and Alcoholism. Unlike normative samples^8,36 where social and enhancement motives are common, participants’ scores on the coping motives scales among the most highly-endorsed. These findings speak to the population—individuals leaving psychiatric care for anxiety and/or depression.

Percentage of days of drinking to cope was significantly associated with the coping and enhancement motives, while days of alcohol use was associated with problematic drinking, coping with depression, and enhancement. Thus, consistent with previous research,⁴² it appears that individuals drink for a multitude of reasons: not only to reduce negative affect but also to improve positive affect. As coping and enhancement are both internally-motivated reasons for drinking (i.e., are aimed at changing or augmenting internal emotional states), this finding is not surprising given the population sampled.

Notably, significant reductions in percent days of use, percent days of using to cope, problematic drinking, coping motives, social motives, and conformity motives were observed over the course of the intervention period. Though this was a small sample, the size of these changes were significant. Perhaps the CHOICE intervention made participants more aware of their alcohol use (particularly their use of the coping motive), thus corresponded with a reduced drinking and drinking to cope rate. However, a future randomized controlled trial and subsequent dismantling studies are necessary to determine the underlying mechanisms of changes.

Project CHOICE appeared to be feasible and accessible. However, some difficulties should be noted. During recruitment, the majority of the participants ruled out for no alcohol consumption or an AUDIT score < 3 (68.4%). As many patients at YAPH were under the age of 21, it is possible they were underreporting their alcohol consumption. It is uncertain if results would generalize to non-clinical populations. Other generalizability difficulties include other substance use: many individuals in the psychiatric program were using cannabis at a high level, either in conjunction with alcohol or reporting no/little use of alcohol (as rule-outs included moderate to severe other substance use and alcohol use on at least 2–3 days/week). Indeed, cannabis use was frequently endorsed in this population. Our previous research in the psychiatric program found that only 34% of emerging adults in the program reported no alcohol and no cannabis use.⁴³ Of the substance users, approximately 80% reported alcohol use and 72% of substance users reported cannabis use. As such, future interventions should emphasize co-use of alcohol and cannabis as well as the coping motive.

Participants often expressed during the follow-up that they had not realized the frequency of which they were drinking to cope prior to their baseline assessment, or the role that emotions played in their alcohol use. All participants reported that they planned to continue to monitor their drinking to cope. Participants also identified with the personalized coping skills messages, and reportedly utilized the messages at least half of the time Thus, it is possible that Project CHOICE helped increase awareness of the participants’ own behaviors and potential negative outcomes, and helped to identify and utilize more adaptive coping skills.

The current study builds on the previous PFIcope intervention and related research in several ways. First, alcohol use is rarely targeted in primarily-psychiatric settings. We targeted the motive of drinking to cope with negative affect among EAs newly-discharged from a psychiatric partial hospital program. Second, the normative feedback portion of the intervention is brief and can be delivered with minimal training in a variety of settings. Third, the current study aimed to implement a PFI pairing EMA with an EMI intervention, triggered in real-time. Fourth, although EMA studies have been conducted to monitor mood, alcohol use, and motive for drinking over time,^31,32 and several EMA/EMI interventions exist for alcohol use and are moderately efficacious,⁴⁴ no known EMI study has specifically targeted a reduction in drinking to cope with negative affect. Fifth, research suggests that EMI effects may be augmented by including face-to-face contact (such as after a brief in-person intervention) and during risky hours (such as in the afternoon before drinking onset). Project CHOICE builds on an in-person PFI: the addition of EMA/EMI to a PFI could effectively train an individual to self-monitor negative affect and coping motives while helping instill/promote adaptive coping strategies.

Several limitations should be noted. This study recruited a small sample. In this open pilot, there was no comparison group. Therefore, we cannot attribute changes in outcomes to the intervention. Given that this was a psychiatric population, findings may not generalize to other young adults who drink. Additionally, mood was not measured at follow-up, thus we were unable to determine if rates of drinking to cope at follow-up was related to mental health symptom reductions. Additionally, the yearly software cost of approximately $3,375 along with programing components may be a burden to providers. Given that there was an in-person component of the intervention, it is unknown if the EMA/ EMI, the in-person session, the partial hospital program, or medication was the driving factor behind the noted effects. Additionally, it is uncertain whether the 6-week, 4x/day EMA schedule was a burden to participants, as evidenced that 40% of eligible participants declined participation and EMA completion rates dropped over time. Furthermore, since only 44% of drinking was motivated by a desire to cope with anxiety and depression, our intervention may only be effective during those times. Despite these limitations, initial feasibility findings are promising.

Project CHOICE may be an ideal intervention for individuals in outpatient care who use alcohol to cope with alcohol and/or depression. It requires minimal financial resources and, once programed, requires no additional clinician time. Additionally, this intervention can be personalized for each individual. The intervention can be delivered in real time, thus Project CHOICE allows for a just-in-time coping strategy when an individual is in a high-risk situation. Given that this project was an open pilot, a randomized controlled trial is warranted to determine the true impact of the personalized coping motive feedback and EMA/EMI.