Pilot Randomized Controlled Trial of a Video Self-Help Intervention for Depression based on Acceptance and Commitment Therapy: Feasibility and Acceptability

Abstract

A common setting where depression is identified and treated is in primary care, where there is a need for low-intensity and cost-effective interventions to be used as part of a stepped-care model. The current study involved a pilot, parallel-group, randomized controlled trial of a video self-help intervention for primary care patients based on Acceptance and Commitment Therapy (ACT). The intervention, called LifeStories, consisted of storytelling vignettes of patients describing their use of ACT-consistent coping skills for depression. Primary care patients were recruited to determine feasibility, acceptability, and potential clinical effects of the intervention. A total of 21 participants were assigned to use LifeStories over a period of 4 weeks, and 19 participants were assigned to an attention-matched comparison group. Qualitative feedback indicated that participants using LifeStories found the intervention to be engaging and useful in transmitting key ACT principles. Furthermore, those receiving LifeStories rated their level of “transportation” or immersion in the videos higher than the control group. Both conditions showed large improvements in levels of depression at a 12-week follow-up. There were no significant differences in symptom outcomes between groups; however, because this was a pilot study, it was not powered to detect differences between interventions. Both conditions additionally showed smaller effect size changes in psychological flexibility, a key ACT mechanism. The results suggest LifeStories to be a feasible and acceptable psychological intervention that may improve depression, and further research is warranted to determine its effectiveness as part of a stepped-care approach to treating depression in primary care.

Introduction

Depression represents one of the leading causes of disability worldwide (World Health Organization, 2017). Although antidepressant medications are a commonly used treatment, surveys of patient attitudes demonstrate that talk therapy is a preferable option for the majority of individuals (Dorow, Löbner, Pabst, Stein, & Riedel-Heller, 2018). However, many barriers exist for individuals wishing to access traditional in-person psychotherapy, including financial burden, geographic restrictions, privacy concerns, and stigmatization (Mohr et al., 2010). Perhaps because of the aforementioned barriers, a frequent setting where depression treatment is first provided is in primary care settings, where up to 10% of patients are diagnosed with major depression (Craven & Bland, 2013), even though the percentage of primary care providers who choose to screen patients for depression continues to be low (Akincigil & Matthews, 2017). Although primary care providers are an important “frontline” in depression treatment due to their familiarity with and access to patients, providing extensive care for depression in these primary care settings is challenging due to time constraints and lack of specialized training (Fleury, Imboua, Aubé, Farand, & Lambert, 2012). One viable option to address these limitations is to provide patients with evidence-based self-help therapies, which in many cases have shown comparable outcomes to in-person interventions when delivered in primary care settings (Cuijpers, Quero, Dowrick, & Arroll, 2019). In particular, internet-based self-help allows psychological interventions to be delivered to patients in an efficient and cost-effective manner, and may be especially useful in treating less severe levels of depression (Newman, Szkodny, Llera, & Przeworski, 2011).

Newer cognitive behavioral therapies (CBTs) such as Acceptance and Commitment Therapy (ACT) (Hayes, Strosahl, & Wilson, 2011) have emerged as efficacious treatments for depression. Compared to traditional CBT, in which treatment is focused primarily on modifying the content of depressive thoughts and related information processing biases, the goal of ACT is to increase psychological flexibility, i.e., the ability to accept internal experiences as they arise in the moment, when doing so is in the service of one’s chosen values (Zettle, 2015). ACT has shown effectiveness in treating depression across a number of studies (Twohig & Levin, 2017), including when adapted for internet-based self-help interventions (Brown, Glendenning, Hoon, & John, 2016). One potential limitation of adapting ACT (as well as other types of CBT) to such formats, however, is the high rate of dropout observed in most self-help interventions. For example, a recent meta-analysis of internet-based depression programs reported that nearly 70% of patients dropped out before completing three-quarters of their assigned intervention (Karyotaki et al., 2015). Thus, while the proliferation of web-based treatments may provide patients much greater access to care, such advances will not have an impact on patients’ clinical outcomes if most do not engage in these interventions.

Narrative communication, or storytelling, may provide a means of delivering therapeutic principles that is more engaging to service users, potentially reducing the risk of dropout observed in traditional self-help. Storytelling is ubiquitous across global cultures (Brown, 2004) and can be defined broadly as the communication of information via “intelligent agents located in a world who participate, or are concerned by, events that change the state of this world” (Ryan, 2017). Whereas traditional self-help formats teach therapeutic principles in a didactic, logic-based style, storytelling facilitates learning that is more experiential in nature (Hinyard & Kreuter, 2007). Narrative approaches may better “transport” the patient into the content being presented so that influence can occur more naturally and effectively (Green & Brock, 2000). Although storytelling has been used for diverse clinical goals, such as treating adolescent mental health (Cook, Taylor, & Silverman, 2004), promoting health-related behaviors (Hinyard & Kreuter, 2007), encouraging cancer screenings (Larkey, Lopez, Minnal, & Gonzalez, 2009), and managing hypertension in minority populations (Houston et al., 2011), we are not aware of any interventions that integrate a storytelling approach with evidence-based principles of change from psychotherapies such as ACT. Given its extensive use of metaphors, stories, and experiential exercises (Foody et al., 2014; Törneke, Luciano, Barnes-Holmes, & Bond, 2015) already, ACT may be particularly well-suited for narrative-based delivery. When combined with the accessibility of web-based self-help, using narrative methods to communicate ACT principles could represent an effective means of targeting depression in broad populations.

Previously, our team developed a video storytelling-based self-help intervention using an iterative, participatory-based research approach to collect personal narratives from community members who had struggled with depression and used effective, ACT-consistent coping strategies as part of their recovery process. In a preliminary open trial (Gaudiano, Davis, Miller, & Uebelacker, 2019), the intervention, called LifeStories, significantly reduced depression symptoms and improved mechanisms targeted by ACT (e.g., mindfulness) in primary care patients over a four-week treatment period relative to a 4-week baseline control period. Additionally, participants rated the intervention as highly acceptable and engaging. In the current study, a pilot randomized controlled trial (RCT), we further examine the acceptability and feasibility of LifeStories in comparison to an attention-matched video comparison condition in order to prepare for a future fully-powered randomized controlled trial. Consistent with current recommended approaches for treatment development research (Kraemer, Mintz, Noda, Tinklenberg, & Yesavage, 2006), the aim of the pilot RCT was to assess the feasibility of the study in a smaller sample before scaling up to a large trial, and not to determine specific treatment efficacy or group differences. By better understanding the feasibility and potential effects of this new innovative narrative-based approach, we hope to further develop an engaging and easy-to-access intervention that addresses the growing public health burden of depression.

Methods

Sample

For study recruitment, brochures were distributed at a large local primary care clinic in addition to online advertisements (e.g., to participate in a study of video-assisted self-help for depression) delivered via Craigslist and Facebook to individuals in the local geographical area. A research assistant completed an initial phone screen with potential participants who contacted us after seeing the study advertisements. If they met initial criteria (described below) assessable over the phone, participants were invited to our research clinic where they completed informed consent and then completed a full intake assessment to determine eligibility.

Study inclusion/exclusion criteria were as follows: (a) 18 years or older; (b) currently receiving care from a primary care provider (at least one appointment in the past 2 years); (c) current or lifetime major depressive disorder as determined by the Structured Clinical Interview for Axis I Disorders based on the Diagnostic and Statistical Manual of Mental Disorders-Fifth Edition (First, Williams, Karg, & Spitzer, 2015) (SCID-5); (d) score of 11 of greater, indicating at least moderate depression symptom severity, on the Quick Inventory of Depression Symptoms-Clinician Version (QIDS-C) (Rush et al., 2003); (e) a score of 9 or higher (indicating moderate depression) on the Patient Health Questionnaire (PHQ-9) (Kroenke, Spitzer, & Williams, 2001); (f) not currently receiving psychotherapy or counseling; (g) not currently receiving psychopharmacological treatment from a mental health specialist (e.g., psychiatrist); (h) if taking an antidepressant medication prescribed by a primary care provider currently, no changes in the preceding 6 weeks; (i) no lifetime diagnosis of a psychotic or bipolar disorder as assessed by the SCID-5; (j) no hazardous alcohol or drug use as assessed by the Alcohol Use Disorders Identification Test and Drug Use Disorders Identification Test (AUDIT/DUDIT) (Babor, Higgins-Biddle, Saunders, & Monteiro, 2001; Berman, Bergman, Palmstierna, & Schlyter, 2005); (k) if pregnant, less than 24 weeks gestation; (l) no suicidal ideation or behavior necessitating immediate safety intervention; and (m) ability to watch a video intervention either online or via a DVD mailed to them.

Procedures

This study was approved by the local IRB. After completing a baseline assessment, participants were randomized to either the intervention condition (LifeStories) or an attention matched comparison group (named WholeDay; described below). Randomization was concealed from study investigators/staff and conducted using a computer program with participants stratified according to depression symptom severity (i.e., QIDS-C < 16 or > 16) and gender (male vs female). Both conditions involved watching videos at home over a 4-week period as described below. In both conditions, a new “episode” (or set of thematically-related videos) was delivered to the participant each week via either an online video stream or a DVD, based on participant preference. Additionally, participants in both groups were given a workbook with supplemental activities to use throughout the intervention period. Participants completed weekly self-report questionnaires of acceptability/satisfaction after each video episode, and interviews and questionnaires of psychological symptoms and functioning at week 4 (posttreatment) and week 12 (follow-up). At posttreatment, participants additionally completed a qualitative interview to elicit feedback about the intervention they had received.

Measures

Acceptability.

The primary aim of the pilot trial was to assess feasibility and acceptability. To measure initial beliefs about expected improvement from treatment, we administered the Credibility and Expectancy Questionnaire (CEQ) (Devilly & Borkovec, 2000), a reliable and valid measure, immediately after participants viewed their first video. In addition, we administered measures of video acceptability and transportation (i.e., engagement) weekly after each video viewing using scales developed in our prior study of LifeStories (citation removed for blind review). Acceptability refers to a viewer’s general sense of satisfaction with a video, whereas transportation captures the extent to which a viewer feels “transported” into the narrative world of the video, a quality theorized to make the message of a narrative more persuasive (Green & Brock, 2000). These measures yielded a mean acceptability score and transportation score, each ranging from 1 = “not at all” to 5 = “very much.” At posttreatment (4 weeks), participants were asked about overall satisfaction with treatment using the validated and reliable Client Satisfaction Questionnaire (CSQ) (Larsen, Attkisson, Hargreaves, & Nguyen, 1979).

Qualitative feedback.

Research staff asked questions at the posttreatment interview related to message (“What was the main point of these videos for you?”), acceptability (“What did you like or not like about any of these videos?”), transportation (“What aspects of the videos were engaging or not engaging?”), and adherence/barriers to completing workbook activities (“Did you complete any of the activities? If not, what got in the way?”).

Diagnoses and clinical outcomes.

We used the SCID-5 (Mood Module and Psychotic Screener; First et al., 2015) at baseline to diagnose mood and psychotic disorders. In addition, outcome measures were collected at baseline, posttreatment, and follow-up. Because the study was not fully-powered to detect differences among treatments, we collected data on clinical outcomes to examine change trajectories in an exploratory fashion to inform future efficacy testing and measure selection. The QIDS-C was administered because it is planned as our primary measure of depression severity in a future trial (lower scores correspond to lower depression severity). The QIDS-C has shown reliability and validity in prior samples of primary care patients (Lamoureux et al., 2010). Study assessors were trained to initial interrater reliability (>.80) on the QIDS-C and SCID-5 with ongoing training during the study to prevent drift. Assessors were not blinded to study conditions due to the pilot nature of the trial. In terms of secondary outcomes we also plan to collect in a future trial, we administered the reliable and valid World Health Organization Disability Assessment Schedule 2.0 (WHODAS) (World Health Organization, 2000), which is a general measure of physical and psychosocial functioning (lower scores mean better functioning). We also collected data using World Health Organization-Quality of Life scale (WHOQOL-BREF) (World Health Organization, 1998), which is a reliable and valid self-report measure quality of life (higher scores mean higher quality of life). For the WHOQOL-BREF, we only examined the psychological and social subscales because they were most relevant to the current study. To measure target mechanisms, we administered the Acceptance and Action Questionnaire-II (AAQ-II) (Bond et al., 2011) to assess psychological flexibility (higher scores correspond to higher flexibility).

Treatment utilization.

We asked additional questions during baseline, posttreatment, and follow-up to assess other treatments received for depression during the study period, including prescribed psychotropic medications, psychiatric hospitalizations, and psychotherapy initiation. Adverse events were also assessed at post-treatment and follow-up.

Treatment Conditions

LifeStories is a video series consisting of four 20-30 minute episodes in which coping strategies are communicated through vignettes of personal narratives, in addition to two psychologist “hosts” who relate the content of stories to essential ACT principles. Episode themes include: 1) cultivating awareness of the transient nature of mood experiences and the benefit of trying out various strategies to cope with depression when it arises; 2) clarifying one’s own values and setting personal goals related to these values; 3) developing a more accepting and nonjudgmental stance towards difficult thoughts and feelings and practicing self-compassion and 4) living more mindfully in the present moment in addition to advice regarding finding professional help for depression (i.e. in-person psychotherapy). For a detailed description of the stories and themes within each episode see Gaudiano et al., 2019. All of the videos within each episode were contained within a single web page that was unlocked by the study participant using a password emailed to them for that week. This ensured that the episodes were viewed sequentially while also allowing participants to re-watch content from prior weeks if desired. Each episode also included “bonus content” where viewers could also learn more about some of the storytellers.

When selecting a comparison condition, we considered several factors. First, we wished to better account for contact time and expectancies for improvement across conditions for a future large-scale trial. We also wanted to find a comparison that matched the mode of treatment delivery of LifeStories, which is video-based and streamed over the internet, as the public is increasingly turning to the internet for health-related information (Freedland et al., 2019). We further wanted to offer a novel intervention option that might be attractive to participants looking for an alternative approach to treating their depression. Therefore, we chose video-based nutrition education as a comparison condition for this study, as it would allow participants to learn about a topic that could improve their general health and well-being, but that was not a psychological intervention directly targeting depressive symptoms. In the future, we expect that LifeStories will be superior to this or similar comparison condition in decreasing depressive symptoms when tested in a fully-powered clinical trial. Our use of nutrition education is similar to the choice to use health education as a comparison group in other studies (Abrantes et al., 2012; Uebelacker et al., 2017). Of course, health education should not necessarily be considered an “inert” intervention. There is an emerging literature suggesting that nutrition and diet interventions could have positive effects on mental health, including depression, even if indirectly (Jacka, 2017; Jacka et al., 2017; Opie, O’Neil, Itsiopoulos, & Jacka, 2015). Ultimately, we chose nutritional education as our comparator because we were not aware of an empirically supported, video-delivered intervention for depression similar to LifeStories that would be a more appropriate alternative in this instance.

WholeDay, the comparison condition created for the current study, was an assembled collection of publicly-available videos related to the development of healthy eating habits. We designed WholeDay as an attention control group that demanded the same amount of time from participants in order to control for the nonspecific effects of watching health-related videos and isolate the key active and specific intervention components (i.e. ACT coping strategies for depression). We selected videos from online streaming sites that were focused on various facets of nutrition, and sent links to these videos to study participants each week, similar to LifeStories. Video topics included: eating a balanced diet, making sustainable healthy lifestyle changes, making informed decisions about food when grocery shopping or dining out, and preparing meals in advance.

Data Analysis

As our primary aim was to study feasibility and acceptability, we focused on our ability to recruit and retain participants into the trial, and perceptions by our participants about the acceptability and credibility of our interventions and randomization procedures. We examined baseline differences between groups (using chi square and t-tests) to assess the outcome of our randomization procedures. All tests alpha were set at p < .05. For descriptive purposes only, we reported effect sizes (intent-to-treat) using Cohen’s d statistics including the corresponding confidence intervals of effects to examine change trajectories and the potential magnitude of effects over time. We used the more conservative last observation carried forward (LOCF) method for handling data that were missing (less than 10% of the sample). Also in an exploratory fashion and to help inform future investigations, we tested for potential differences between conditions over time using Analyses of Covariance (ANCOVAs) accounting for baseline scores.

Responses to open-ended qualitative questions were audiorecorded and transcribed. One author (LU) developed an initial qualitative coding scheme that included codes for main themes of videos, names of the various storytellers, and response to the workbook. Through the coding of transcripts, the team added further codes to the codebook. Initially, one author (LU) coded all transcripts; a second author (BG) then read through all transcripts, and made suggestions for additional coding. Any differences were resolved via consensus. LU then wrote descriptions summarizing each code and chose exemplar quotes. CD reviewed these summaries and quotes and provided additional suggestions. Again, all differences were resolved via consensus. LU then divided codes into five larger categories. In the Results section, all codes are denoted with italics, with study participant number for particular quotes in parentheses.

Results

Sample Characteristics

A total of 185 people completed the initial phone screen, 141 did not meet study inclusion criteria, and 40 were enrolled and randomized to conditions (see Figure 1 for subject flow diagram). Table 1 presents detailed demographic and clinical characteristics at baseline. The final sample was composed of 36 women (90% women), with 33 participants identifying as White (82.5%). In addition, 7 participants (17.5%) identified as Hispanic/Latino. A total of 16 participants (40%) were currently married or cohabitating, and 18 (45%) were employed full/part-time. The sample had a mean age of 47.8 years (SD = 13.2) and 20 participants (50%) reported earning a college degree or higher.

Figure 1.

Diagram of participant flow

Table 1.

Baseline participant demographics and clinical characteristics

	LifeStories (n=21)		WholeDay (n=19)
Demographics	n or mean	% or SD	n or mean	% or SD
Gender
Male	3	14.3%	1	5.3%
Female	18	85.7%	18	94.7%
Agea	46.9	12.2	48.8	14.5
Race
White or Caucasian	16	76.2%	17	89.5%
Black or African American	2	9.5%	0	0%
Other or Multiracial	3	14.3%	0	0%
Ethnicity
Hispanic/Latino	4	19%	3	15.8%
Not Hispanic/Latino	16	76.2%	14	73.7%
Marital Status
Married/Cohabitating	10	47.6%	6	31.6%
Single/Divorced/Separated/Widowed	11	52.4%	13	68.4%
Education
Less than High School	0	0%	1	5.3%
High School/GED/Some College	11	52.4%	7	36.8%
College/Master’s Degree	10	47.6%	10	52.6%
Employment Status
Employed Full- or Part-time	10	47,6%	8	42.1%
Student/Homemaker/Retired	6	28.6%	2	10.5%
Unemployed	2	9.5%	2	10.5%
Disability	3	14.3%	6	31.6%
Income
$0-25,000	4	19%	6	31.6%
$25,000-50,000	6	28.6%	7	36.8%
$50,000 or greater	10	47.6%	5	26.3%
Clinical Characteristics
Current Major Depressive Disorderb	20	95.2%	16	84.2%
Persistent Depressive Disorderc	11	52.4%	8	42.1%
Baseline Depression Severity (QIDS-C)d
Moderate Range	11	52.4%	12	63.2%
Severe Range	10	47.6%	6	31.6%
Very Severe Range	0	0%	1	5.3%
Currently Taking an Antidepressant	10	47.6%	9	47.4%

Note. Some demographic variables do not total 100% because not all participants provided answers.

^aContinuous variable

^bMDE = Major Depressive Episode, as assessed by the Structured Clinical Interview for DSM-5 (SCID-5)

^cParticipants with chronic depression meet criteria for persistent depressive disorder as assessed by the SCID-5

^dQIDS-C = Quick Inventory of Depression Symptoms- Clinician Version; no participants were in the “None” or “Mild” severity ranges on the QIDS-C

In terms of diagnoses, 36 participants (90%) met DSM-5 criteria for a current major depressive disorder, with 19 (47.5%) also meeting criteria for current persistent depressive disorder. According to the QIDS-C, 23 participants (57.5%) were in the moderate range for depressive symptoms, 16 (40%) were in the severe range, and one (2.5%) was in the very severe range at the beginning of the study. At baseline, 19 participants (47.5%) were taking an antidepressant prescribed by their primary care provider.

A total of 21 participants were randomized to the LifeStories condition, and 19 were randomized to the WholeDay condition. To examine potential differences between groups on baseline variables, we conducted t-tests and chi-square tests. No significant differences were found between groups on baseline variables (ps > .10). However, visual examination of characteristics suggests that the LifeStories group was somewhat more severe in terms of initial depression severity based on higher rates of current major depressive disorder and persistent depressive disorder (see Table 1).

Treatment Utilization and Safety

At posttreatment and follow-up, 11 participants (52.4%) in the LifeStories group and 8 (42.1%) in WholeDay were taking an antidepressant, χ² = 0.42, df = 1, p = .52. Concurrent psychotherapy was an exclusion criterion at baseline but some participants started therapy during the trial. One participant (5.3%) from the WholeDay group had started seeing a therapist by posttreatment. At week 12 follow-up, two participants (9.5%) in LifeStories and five total (26.3%) in WholeDay had begun psychotherapy, χ² = 2.33, df = 1, p = .13.

Study staff asked about adverse events at each study assessment. No adverse events were reported that were unexpected or related to study participation. No participants were psychiatrically hospitalized, had attempted suicide, or were clinically deteriorated (defined as an increase in QIDS-C score of 6 or more points, and a total score greater than or equal to 16) through the follow-up period, suggesting that both interventions were generally safe.

Feasibility and Acceptability

From the 40 participants who completed baseline and were randomized to conditions, 37 completed the posttreatment assessment (at 4 weeks), and 36 completed the follow-up assessment (at 12 weeks) (see Figure 1). By 4-week post-treatment, no participants had dropped out of LifeStories and one participant dropped out of WholeDay due to reportedly being too busy to complete study tasks. There were no significant differences between proportion of dropouts in LifeStories (4.8%) vs WholeDay (15.8%), χ² = 1.35, df = 1, p = .25.

See Table 2 for acceptability and satisfaction outcomes between groups. Participants displayed overall high levels of adherence to the interventions, with those in the LifeStories and WholeDay groups watching an average of over 3 videos each. We collected the CEQ to assess beliefs related to the credibility of the interventions after participants viewed the first video in each condition. Although not statistically different, the mean CEQ Total scores were 31.4 in the LifeStories group and 27.1 in the WholeDay group, out of a possible score of 54 total. When rating overall program satisfaction at posttreatment, those in the LifeStories group had an average CSQ-8 rating of 23.5 and those in the WholeDay group had an average rating of 22.5, out of a possible score of 32 total, which was not statistically different.

Table 2.

Descriptive statistics for treatment feasibility/acceptability measures and between-group effect size differences

	M (SD)		t (df)	p	Cohen’s d effect size (between subjects) with 95% CI
Measure	LifeStories n = 21	WholeDay n = 19
Acceptability Average	4.3 (0.4)	4.2 (0.5)	1.09 (34)	.283	0.22 [−0.44, 0.87]
Transportation Average	3.5 (0.5)	3.2 (0.3)	2.47 (34)	.019*	0.73 [0.04, 1.39]
CEQ	31.4 (10.0)	27.1 (9.5)	1.36 (36)	.181	0.44 [−0.21, 1.08]
CSQ-8	23.5 (3.4)	22.5 (3.1)	0.89 (32)	.378	0.31 [−0.38, 0.98]
No. Videos Viewed	3.43 (1.21)	3.79 (0.92)	−1.06 (38)	.299	−0.33 [−0.95, 0.30]

Note. CI = confidence interval; CEQ = Consumer Expectancy Questionnaire; CSQ = Client Satisfaction Questionnaire

^*p < .05

In addition, participants rated the acceptability and “transportation” for each video after viewing. Average ratings of video acceptability were found to be high and non-significantly different across conditions. However, a significant difference was observed between the average video transportation ratings across all episodes by condition, favoring the LifeStories group (see Table 2).

Qualitative Results

Category 1: Planned video themes.

Participants responded in ways that indicated they understood important themes that were presented in LifeStories. Reflecting material introduced in episode 1, five participants commented on the transient nature of thoughts and feelings, including the importance of remembering that these are transient. One participant commented: “I wrote, or reinforced that thoughts and feelings are constantly changing…it’s nice to remember that as a tool when you feel like the bottom is falling out from underneath you” (4166). Participants also commented on the theme of taking small behavioral steps toward a goal (n=6). They mentioned meaningful activities that they had engaged in, and one mentioned using a calendar (discussed in one of the videos) as a way to track activities.

Consistent with material introduced in episode 2, participants identified the importance of identifying values and living a life consistent with values (n=6). For example, a participant stated: “Live a life that you can be proud of. Live intentionally….I felt guilty smoking weed, you know what I mean? So, then let’s find another way, let’s find a way that doesn’t make you feel guilty, and that stood out for me. That was huge for me….I’m not smoking weed anymore” (4268).

Participants also commented on themes introduced in episode 3. With regard to emotional awareness and acceptance, participants reported understanding the importance of identifying, expressing, and accepting their feelings as a result of watching the videos (n=8). One participant commented: “all you need to do is to be able to be there and experience the feeling and give it the time you need to get through it. When you get through it there is going to be some kind of new awareness about the process of getting through it, even the dirty parts of life.” (4116). Four participants commented on developing self-compassion being an important message in LifeStories. For example, one person said: “the thing I took home was that it’s okay to feel this way and not to beat yourself up over it” (4212).

Finally, as discussed in episode 4, participants commented on the theme of mindfulness, meditation, and breathing practices as ways to cope with depression (n=7). They were largely interested in interested in trying more meditation and breathing exercises, though two people noted that there are difficulties in practicing meditation, including “screaming kids” (4114) and self-discipline. Participants recognized the theme of engaging in therapy or getting professional help (n=7). Specific comments on therapy varied, and included their own need for therapy; telling their story to a new therapist being a barrier to therapy; the idea that even if one therapist doesn’t work out, another might; and feeling comforted in knowing that there is help available if needed.

Category 2: Other video themes.

Over three-quarters of participants commented on the positive aspects of knowing that others experience depression too, and that those people were able to get through problems using various coping strategies (n=16). For example, in response to the question “What did you learn from this video?” one participant stated: “That I am not alone. There are a lot of people that have similar issues and have had similar things happen to them in their lives. And they succeeded.” Other themes that participants reported taking away from the videos included focus on the positive, i.e., focus on something one enjoys instead of always focusing on depression (n=3), and focusing on the present moment instead of the past or future (n=3).

Category 3: Storytellers and narrators.

Different participants reacted in different ways to the various storytellers. In many cases, there were participants who really related to a particular storyteller (n=14), often because the participant could relate to a particular similar life experience, such as having a parent with dementia, having gotten divorced, having a history of abuse, or interpersonal style. One participant said: “I just liked the fact that it was real people I could relate to in my area, and that they were so open and honest. And also that it was a variety of circumstances… So I could relate” (4162). Another stated: “it was almost like I was talking to one of my friends” (4055). At the same time, participants commented on difficulty relating to storytellers (n=6) with different life experiences, such as having problems with anger, being abused as a child, or having a substance use history. For almost every storyteller with whom someone could not relate, others commented that that they could relate well to that storyteller. Two people commented that they liked that there were a variety of storytellers with different life experiences.

Finally, most participants did not specifically comment on the narrators, but the three people who did provide responses had mixed evaluations. One participant stated: “But when you have the clinicians on, there’s nothing bad about that, but I think everybody recoils a little bit from the clinical tone… But then you jump into the stories and you go, okay, this is a lot easier for me and it’s no longer abstract” (4171). Another person liked listening to the narrators, and a third commented that they did not like the style of going back and forth between the two narrators in the videos.

Category 4: Overall reactions to videos.

In response to the question “What did you not like about this video?” Eleven participants said there was nothing they did not like. Six people had trouble remembering the videos at the time of the interview: 2 people did not remember details but clearly remembered key messages; 1 remembered one key message; and 3 remembered very little. With regard to the videos being boring (n=2), one participant stated: “some of the stories are nice to hear, cause it shows… that you are not alone. But, I find them kind of boring….I like to watch action and adventure movies” (4225). Another stated: “They weren’t the most exciting videos in the world. But they weren’t terribly boring either” (4212). Three participants stated that nothing about the videos was strongly engaging; one did not watch beyond the first video.

Category 5: Workbook.

Responses to the workbook were mixed. Ten people used the workbook and found it helpful. For example, a participant stated: “…instead of writing in the workbook I actually made copies of the sheets. So that if I’m having trouble figuring out how I’m feeling throughout the day, I can fill out the emotion log more than once.” (4198). However, six people did not use the workbook, and two other participants stated that they did not like writing things down or having “homework” – although one of those two people stated that they do write in their journal.

Preliminary Clinical Outcomes

See Table 3 for descriptive statistics and within-subject effect size estimates for outcome measures in each condition. On the primary depression measure (QIDS-C), large effect size changes were observed in both the LifeStories and WholeDay conditions by follow-up. In contrast, small to medium effect sizes improvements were generally observed for each group on secondary outcomes (quality of life and functioning) and our process measure (psychological flexibility). A series of 2 (LifeStories vs WholeDay) x 2 (Post-treatment and Follow-up) ANCOVAs with baseline scores covaried, revealed no significant differences between groups over time (all ps = n.s.). Confidence intervals for all outcomes were large and overlapping between the conditions as expected given the small sample size. Therefore, no definitive conclusions or generalizations can be drawn about potential groups differences as these effects would not necessarily reliability generalize to larger samples (Kraemer et al., 2006).

Table 3.

Descriptive statistics for outcome measures and within-subjects effect sizes for intent-to-treat sample (n = 40)

	M (SD)			Cohen’s d effect size change with 95% CI
Measure	Pre	Post	Follow-Up	Pre to Post	Pre to Follow-Up
Primary Outcome
QIDS-C
LifeStories	15.0 (2.0)	10.2 (3.9)	9.0 (4.8)	1.55 [0.83, 2.21]	1.63 [0.91, 2.30]
WholeDay	15.0 (2.5)	8.0 (3.5)	7.8 (4.3)	2.30 [1.44, 3.07]	2.05 [1.22, 2.78]
Secondary Outcomes
WHO-DAS
LifeStories	12.8 (6.1)	13.1 (7.1)	11.3 (5.4)	−0.05 [−0.65, 0.56]	0.26 [−0.35, 0.86]
WholeDay	15.7 (7.7)	11.9 (6.2)	12.1 (8.5)	0.54 [−0.11, 1.18]	0.44 [−0.21, 1.08]
WHOQOL-BREF-Psychological
LifeStories	21.0 (6.5)	21.2 (8.0)	25.0 (9.2)	−0.03 [−0.63, 0.58]	−0.50 [−1.11, 0.12]
WholeDay	18.3 (7.4)	24.4 (10.0)	23.4 (11.0)	−0.69 [−1.33. −0.02]	−0.54 [−1.18, 0.11]
WHOQOL-BREF-Social
LifeStories	60.5 (26.5)	65.2 (22.6)	69.0 (24.8)	−0.19 [−0.79, 0.42]	−0.33 [−0.93, 0.28]
WholeDay	64.8 (23.0)	66.9 (20.2)	64.2 (19.8)	−0.10 [−0.73, 0.54]	0.03 [−0.61, 0.66]
Process Measure
AAQ-II
LifeStories	29.5 (9.6)	28.3 (8.9)	29.8 (7.2)	0.13 [−0.48, 0.73]	−0.04 [−0.64, 0.57]
WholeDay	30.4 (8.8)	29.4 (7.5)	32.8 (8.2)	0.12 [−0.52, 0.76]	−0.28 [−0.92, 0.36]

Note. LifeStories n = 21; WholeDay n = 18; CI = confidence interval; QIDS-C = Quick Inventory of Depression Symptoms; WHO-DAS = World Health Organization Disability Assessment Schedule; WHOQOL-BREF = World Health Organization Quality of Life scale; AAQ-II = Acceptance and Action Questionnaire

Discussion

Results from this pilot RCT of ACT-based storytelling videos versus nutrition education videos for depression indicated that both conditions were deemed safe, feasible, and acceptable to participants. Most of the study sample met criteria for current major depression and demonstrated a range of depression severity, suggesting that this was a clinically relevant group. We were able to meet our recruitment targets, we witnessed very little drop out during the trial and follow-up period, and most participants completed the videos as intended. Importantly, no adverse events were reported during the study. Furthermore, participants rated the videos as credible and acceptable after viewing, and overall they were satisfied by the interventions at post-treatment. However, the LifeStories intervention was rated significantly higher on “transportation” or immersion with content compared with the nutrition condition, which was an important treatment goal based on the narrative intervention format. As discussed, the study was not powered to detect differences between conditions. However, broadly speaking, both interventions were associated with similarly large improvements in depression severity through follow-up. In contrast, the changes observed on quality of life and functioning measures were only in the medium range, and thus future studies are needed to verify effects on these constructs. Furthermore, small effect size changes were witnessed in both groups on our ACT process measure of psychological flexibility. It is possible that a sufficient dose or intensity of the intervention was not delivered to move this construct. While this treatment development study did not elucidate significant differences between conditions, qualitative findings (see below) suggest that video-based psychosocial treatments such as LifeStories have the potential to effectively communicate evidence-based therapeutic principles, and thus may show advantages over comparison treatments (such as WholeDay) when tested at a larger scale in terms of their feasibility and relevance to the clinical population. Overall, the study met its intended treatment development targets, and thus the next step is to test of the effectiveness of LifeStories in a future full-scale clinical trial to determine efficacy.

Qualitative feedback regarding participants’ experience with LifeStories documented that most understood the key themes or messages of the intervention, which included: recognizing the transient nature of thoughts and feelings, making small behavioral changes toward goals, identifying and living consistently with personal values, being more emotionally aware and accepting, fostering self-compassion, practicing mindfulness, and engaging in formal therapy when needed. Generally, participants found the storytellers featured in the videos to be credible, engaging, and relatable; although there was more mixed feedback as to the usefulness of the professional narrators. This feedback suggests that some re-editing of the videos may be indicated to focus more on the storytellers and less on the professional guidance. There also was mixed feedback regarding the utility of the accompanying workbook, with some participants finding it a helpful adjunct to the videos, whereas others failed to use it. Thus, modifications to the workbook might be helpful to increase engagement. In future iterations of LifeStories, one option would be to use an interactive, online format for the self-help exercises that accompany of the videos, rather than a printed workbook.

Several limitations should be considered when interpreting findings. First, the study was not powered to identify differences between conditions, and thus a fully-powered RCT is needed to verify the effectiveness of LifeStories. The large confidence intervals around effects suggest that any numerical differences observed between conditions may not be reliable for predicting effects in future samples. Second, most of the sample was comprised of individuals who identified as female, and further study of LifeStories is needed in a more gender diverse sample. Third, the study contained participants with a range of depression severity, but the sample size was too small to investigate subgroup differences or moderators of effects. Fourth, LifeStories is designed as a minimally-intensive intervention delivered over a relatively brief time period (one month), and the study follow-up period was only 12 weeks total. Therefore, it is unclear how well the effects of the intervention on depression severity were maintained in the longer term. In addition, participants were engaged in other traditional treatments during the trial (e.g., antidepressant medication and psychotherapy), which could have influenced clinical outcomes. Finally, our comparison group consisted of health and nutrition videos, and thus LifeStories was not compared with a validated self-help intervention for depression. Therefore, future research could also compare LifeStories with other more traditional self-help interventions to understand advantages and disadvantages (e.g., bibliotherapy or online interventions). We would expect LifeStories to show equivalency for treating depression when compared with other validated self-help approaches, and thus it could offer an alternative choice for patients based on their preferences. We also anticipate that LifeStories could demonstrate superiority to other self-help approaches in other ways, such as in engagement or adherence, given its narrative video format.

Overall, results of the current pilot trial were encouraging and indicate the need for future studies of LifeStories. Participants found the stories from the individuals with lived experience recovering from depression to be engaging and useful for conveying important messages key to ACT. Ultimately, we do not view LifeStories as a standalone treatment for depression, but rather, to be used as part of a larger stepped-care model. Such stepped-care models start with low-intensity, low-cost treatments and then move to higher intensity interventions only when needed. Typically, the initial step is “watchful waiting,” followed by guided self-help, preferably delivered via the internet or mobile technology, and then by traditional evidence-based psychotherapy, with psychotherapy plus antidepressant medication as the final step for more severe or chronic cases (van Straten, Seekles, van’t Veer-Tazelaar, Beekman, & Cuijpers, 2010). The relevance of interventions such as LifeStories is particularly clear in primary care settings, where physicians may be able to identify that depression is a problem, but are limited in terms of time and resources as to how to respond effectively and efficiently (Gum, Epstein-Lubow, Gaudiano, Wittink, & Horvath, 2019). If ultimately found to be safe and efficacious in larger-scale studies, LifeStories could play a useful role as part of a stepped-care approach for treating depression in primary care due to its low-cost and ease of delivery.

Key Practitioner Message.

• Narrative-based self-help interventions such as LifeStories have the potential to communicate principles from evidence-based therapies.

• Self-guided resources for managing depression could be used as one component of a primary care stepped-care model in which intervention intensity is matched to patients’ care needs.

• Online interventions such as LifeStories provide a potentially accessible and scalable means of providing care to patients with depression for whom treatment access and costs are barriers.