A randomized controlled trial of a web-based personalized feedback intervention targeting frequent indoor tanning bed users: Engagement, acceptability, and preliminary behavioral outcomes

Abstract

Frequent indoor tanning bed use is an established public health concern, yet research on tanning cessation interventions for frequent tanners is lacking. We describe the protocol for a brief, web-based tanning behavior change intervention and present evidence that it is acceptable and engaging to frequent indoor tanners. Lower tanning rates were not observed among participants receiving the intervention in a randomized controlled trial but participants’ interest in changing tanning increased. This intervention could be a useful approach to increasing frequent tanners’ interest in behavior change and openness to engaging within a more intensive, multi-component tanning cessation program.

Introduction

The use of indoor tanning beds that emit artificial ultraviolet radiation (UVR) is estimated to cause 400,000 cases of skin cancer in the United States each year and contribute to 1 in 10 of all new melanoma cases (Wehner et al., 2012; Wehner et al., 2014). Melanoma has become the 5th most common cancer in the U.S and the 3rd most common cancer among young adult women, who are the most frequent users of tanning beds (Barr et al., 2016). Although rates of any past year indoor tanning among U.S. adults have declined in the past decade from 1 in 7 in 2007 to 1 in 25 in 2018 (Geller, 2018), the rate of frequent use has increased with 24% of recent tanning bed users reporting tanning 25 or more times in the past year in 2018 compared to 13% in 2007 (Bowers et al., 2020). Compelling evidence from biological and psychological studies show that UVR exposure can have physically reinforcing effects and leads to the concern that excessive tanning may be a result of sensitivity to these addictive properties (Heckman et al., 2016; Noar et al., 2014; Stapleton et al., 2017a). Indoor tanning bed users believe that tanning enhances physical attractiveness and report psychologically-reinforcing feelings of increased confidence and social acceptance from tanning (Noar et al., 2014). Frequent indoor tanning leads to problematic behaviors that are central to models of behavioral addiction (e.g., spending time and resources on tanning that leads to neglecting other responsibilities) (Ashrafioun and Bonar, 2014; Stapleton et al., 2017a). Despite the increasing public health importance of frequent indoor tanning, there is a dearth of interventions designed to target this group, which is an important gap in the skin cancer intervention literature (Stapleton et al., 2017a).

Our intervention is grounded in behavior change concepts derived from the general Motivational Interviewing (MI) framework for interventions intended to motivate individuals to change addictive behaviors (Miller and Rollnick, 2012). Individuals who engage in risky behaviors like tanning experience a natural state of behavioral ambivalence, which is defined as holding both positive and negative views of the behavior. Interventions are designed to promote self-exploration of ambivalence by encouraging individuals to reflect on both the positive and problematic aspects of behaviors. This exploration is intended to make the negative aspects of behavior more salient and produce an increased openness to changing behavior, the first step in the behavior change process. Among individuals who are open to changing, it is important to provide behavior change strategies to boost self-efficacy and likelihood of making a successful change (Miller and Rollnick, 2012). Studies show indoor tanners perceive aspects of their tanning to be problematic, perceive advantages to reducing their tanning, and are receptive or actively trying to change their tanning (Banerjee et al., 2014; Glanz et al., 2018; Harrington et al., 2011; Mosher and Danoff-Burg, 2010). This evidence that some indoor tanners experience behavioral ambivalence and report an openness to changing their tanning behavior suggests an MI framework may be appropriate.

One of the more successful applications of MI behavior change concepts with young adults can be found among brief interventions developed to reduce high-risk alcohol use. Traditionally, these interventions adopt an in-person counseling approach to engage participants in discussions about their alcohol use, beliefs, and related problems and help participants identify strategies for behavior change and reducing harm (e.g., moderate drinking/drinking less) (Carey et al., 2007; Carey et al., 2009; Carey et al., 2012; Cronce and Larimer, 2011; Foxcroft et al., 2014; Huh et al., 2015; Larimer and Cronce, 2007; Miller et al., 2013; Murphy et al., 2010; Scott-Sheldon et al., 2014). More recently, researchers have begun to test alternative intervention delivery modalities to eliminate the need for resource-intensive counselors. Personalized feedback interventions (PFIs) are web-delivered interventions that utilize one-time assessments to gather data regarding users’ alcohol use and beliefs which is used to subsequently produce tailored feedback about an individual’s own drinking patterns, alcohol-related concerns, and strategies to reduce harm (Leeman et al., 2015; Miller et al., 2013; Paulus et al., 2020; Ray et al., 2014; Walters and Neighbors, 2005). To the extent that such feedback elicits reflection on an individuals’ problems stemming from their alcohol use, they may increase their openness to changing their drinking by enacting the provided change strategies.

The current study describes the formative testing of a web-based indoor tanning intervention that adapts the general PFI approach of assessment and tailored feedback related to users’ tanning beliefs and behavior. This novel tanning intervention approach is designed to better address the unique psychological motivations for tanning among frequent indoor tanners and utilizes a behavior change framework that has been shown to be successful in changing addictive behaviors. Our study also introduces some innovations to the typical PFI approach. Rather than deliver feedback after the initial assessment, our intervention provides feedback in real-time as users progress through the assessment. This approach is intended to produce more meaningful reflection of intervention content. We also supplemented the web intervention, designed primarily to motivate openness to change, with a series of four weekly booster interventions designed to encourage change planning, behavior monitoring, and building self-efficacy, all critical skills in successful behavior change. These modifications were designed to offer an experience that more closely aligns with a traditional, in-person delivery approach.

In this study, we report results from a randomized controlled trial of the intervention among a sample of young adult women engaged in frequent indoor tanning. Our study was designed to examine intervention acceptability and engagement. We also assess preliminary behavioral outcomes by comparing the tanning bed behaviors of intervention recipients to the waitlist-controlled condition. Finally, we explore putative mediators of the intervention by evaluating intermediate intervention outcomes consistent with PFI behavior change principles including readiness to change and self-efficacy.

METHODS

Participants and Recruitment

Study eligibility criteria included women between 18 – 25 years of age engaged in frequent tanning bed use, which was defined as at least 25 indoor tanning sessions in the past 12 months (referred to as frequent indoor tanners). Participants were recruited using multiple approaches. Qualtrics Sample Providers emailed study advertisements with links to a brief online eligibility screening survey to members of internet panels and posted online advertisements. The screening survey was programmed to present eligible participants with a brief description of the intervention trial and a question to indicate their interest in receiving an email study invitation. Second, recruitment flyers with study personnel contact information were posted on campus and in classrooms at a large Northeastern University and on internet ads on social media and Craigslist. Most participants (96%) were recruited by Qualtrics.

Study Design and Procedure

The study was a 2-arm randomized controlled trial with planned 1:1 allocation. A study coordinator enrolled participants, administered all procedures, and monitored trial progress. Following screening, the study coordinator sent an invitation email with a link to the online baseline survey, programmed using Qualtrics survey software, and a unique study personal identification number (PIN) for accessing the survey. Each PIN was assigned to a study condition, and we used a random number generator to randomly order the PINs in our study tracking spreadsheet. PINs were assigned to participants in order of our receipt of their screening survey and prior to sending the baseline survey invitation. Participants were asked to provide online informed consent prior to beginning the baseline survey (implied, non-written consent) and those who did not were not provided access to the survey. Up to three email reminders were sent to individuals who agreed to participate. Links to the web-based intervention were sent to participants two weeks after completing the baseline survey. A brief intervention evaluation survey was presented immediately after participants completed viewing the web-based intervention component. Following the web-based intervention, participants received 4 weekly web-delivered boosters designed to be completed in 5-minutes. All participants who completed the baseline were invited to complete the follow-up assessment 12 weeks later. Participants assigned to the waitlist control were invited to access the intervention after the follow-up surveys concluded. Participants received a gift card for completing the baseline survey ($40), the intervention evaluation ($20), and the follow-up survey ($40). Baseline assessments were completed between April 2018 and November 2018, and follow-up assessments were completed between July 2018 and February 2019. The University of Kentucky’s Institutional Review Board reviewed and approved all study procedures as a minimal risk study (protocol approval number 60965) and the trial was registered on clinicaltrials.gov (NCT03448224).

Intervention

The intervention was programmed with Snap survey software (SNAP Surveys Ltd, 2012). Each page was designed to be completed in a pre-determined order and the intervention was designed to take 30 minutes or less to complete and each booster was designed to be completed in 5-minutes. The intervention spanned a total of 29 distinct web pages with 62% containing primarily assessment questions (examples included as Appendix II, III, & V) and 31% containing personalized feedback (examples are Appendix I & IV) (2 pages included introductory and ending information). The intervention was primarily text-based given the restrictions of the survey software platform, with the exception of the graphic feedback for consequences shown in Appendix I. Booster content engaged participants in monitoring their recent tanning behavior and provided additional skills-based behavior change content.

Intervention content development was informed by principles of MI and designed to retain features of web-based PFIs applied to other content areas (e.g., Ray et al., 2014). Content was designed with three main goals (Table 1). First, intervention content was designed to enhance readiness to change by exploring the natural state of ambivalence of risk behaviors, defined as holding both positive and negative views of the behavior (Miller and Rollnick, 2012). The assessment of tanning behavior and experiences combined with the provision of personalized feedback is designed to encourage self-reflection of risk and risky behaviors and foster exploration of tanning ambivalence. Users answered questions about their individual patterns of indoor tanning bed use, phenotypic skin features (e.g., skin color, freckles) and other skin cancer risk factors, consequences they experienced related to tanning (e.g., getting into disagreements about tanning, neglecting responsibilities to go tanning), and tanning-related problems that represent symptoms of behavioral addiction to tanning (e.g., feeling guilty about tanning, urges to tan, attempts to control tanning). Responses were used to create personalized feedback on their annual amount of tanning behavior, including: the number of standard erythemal doses (i.e., a measure of the typical amount of sun exposure individuals receive in a year) they received from tanning beds in the prior year, risk of developing skin cancer and related problems, a comparison of the participant’s level of tanning-related consequences as compared with other tanners based on data from a national survey of indoor tanners (Stapleton et al., 2017b) (see Appendix I as an example of this feedback), and a summary of the disruptiveness of tanning in their daily life. Subsequent intervention screens contained several questions about whether participants agreed with statements of various perceived benefits and costs of tanning. Participants were then asked to consider the balance of the benefit to cost comparison (in the context of their important values, additional description below).

Table 1.

Intervention map with the goals and brief descriptions of intervention content.

Goal	Content and Description
Enhance interest and readiness to change	Self-exploration of tanning ambivalence. Participants were encouraged to answer open and close-ended questions in which they described individual patterns of indoor tanning bed use, experiences with tanning, and their views of the of pros (“good things about tanning”) and cons (“not-so-good things about tanning”) of indoor tanning. Subsequent personalized feedback was tailored to present the pros and cons that a user selected and users were asked to indicate their perception of the balance of pros vs cons.
Bolster commitment to change	Normative beliefs about IT. Participants were asked to estimate the percent of women their age who use IT. Following this exercise, participants were provided with the actual percent of women compared to what they believed to be true.
	Desired motives and values. Tanners were asked questions designed to clarify their most important values or priorities. The values and descriptions presented were:  - Health and fitness: Being physically healthy, fit, and strong  - Relationships: Having deep, enduring relationships with family, friends, or someone I love. Fitting in and having a strong group of friends.  - Image: Looking good and having an attractive appearance. Achieve the look I’ve been after.  - Self-esteem and well-being: Knowing and accepting who I really am. Experiencing personal growth. Feeling relaxed, having good moods, and feeling less stressed. Through a series of close-ended questions, participants were then encouraged to think about the good and not-so-good aspects of indoor tanning. Following this exercise, personalized feedback was provided to the participant aligning their personal values with the pros and cons of indoor tanning.
Encourage change planning	Skills necessary for successful behavior change. Information about what other tanners believed to be the benefits of changing their indoor tanning behavior was provided to participants. Two alternatives to indoor tanning were reviewed and the participants evaluated whether or not the alternative could enhance their physical appearance, boost their mood, damage their appearance, or cause health problems.
	Examples of strategies for change. A personalized change plan was shown to the participant offering new procedures for meeting the participants tanning needs in a healthier way. This included a strategy and detailed instructions on how to reduce tanning-related harm. The change planning process concluded with the participant evaluating their own thoughts about changing their tanning. Four weekly boosters designed to assist the participant in monitoring their tanning behavior and adherence to their change goals were emailed to participants.

The second goal was to bolster participants’ commitment to change by identifying perceived discrepancies with current tanning behavior and important external or internal standards (Neal and Carey, 2004). External standards were addressed with a personalized normative feedback approach designed to correct normative misperceptions about tanning that result from the tendency of frequent indoor tanners to overestimate the prevalence of tanning behaviors of peers (Carcioppolo et al., 2019). Intervention users were asked to provide estimates of the percentage of young women who used any indoor tanning in the past year, tanned 10 or more times, and tanned 25 or more times and subsequent feedback showed how these estimates compared to the actual percent of young adult women who engaged in tanning bed use at these levels based on national survey data collected near the time of intervention development (Stapleton et al., 2017b). Providing data to demonstrate these normative discrepancies is consistent with MI principles and has been shown in experimental research to reduce tanning intentions among frequent users (Carcioppolo et al., 2019).

To address discrepancies between tanning and internal values, participants were presented with four key values or life priorities (i.e., health & fitness; relationships; image; self-esteem & well-being) and rated how important each value was to them. Users were then presented with multiple sets of statements of various perceived benefits and costs of tanning and indicated their level of agreement with each statement. These benefits and costs were selected based on existing literature and corresponded to one of the four personal values. For example, Appendix II shows questions related to the cons of tanning for relationships and well-being. For the personalized feedback portion, a subsequent intervention screen displayed a listing of the pros and cons relevant to each value that were endorsed in prior screens by the user. Users were asked to indicate how they viewed the balance between pros and cons for each value (Appendix III). To the extent that reflection on the undesirable aspect of tanning fosters a shift in the balance of pros and cons in tanning expectancies or creates a perceived misalignment with values, commitment to changing tanning should be strengthened (McNally and Palfai, 2003; Neal and Carey, 2004).

The final intervention goal was to encourage change planning and the consideration of alternative ways to obtain the benefits of tanning while reducing the risks. First, participants were presented with reasons for changing their tanning that were identified in our prior qualitative research (e.g., avoid skin cancer, save money, less worry, healthier skin) and indicated the importance of making a change in their tanning. Participants were then asked to evaluate how sunless alternatives (i.e., spray/airbrush tanning and sunless tanning lotions) compared to indoor tanning on several factors (e.g., enhancing physical appearance, possibility of causing health problems). Participants were presented with multiple suggestions for reducing their tanning (e.g., decide to stop tanning entirely; space out the times between tanning sessions) and reducing temptations to use tanning (Appendix IV). Strategies for reducing tanning harm were included along with a link to a Facebook page with information about sunless tanning products. The series of four weekly booster interventions following the intervention asked participants to report on their past week tanning behavior, monitor any progress they made on their change goals, and identify new strategies for reducing their tanning.

Measures

Acceptability.

Consistent with our prior intervention trials, intervention acceptability was measured using four general intervention evaluation items (Hillhouse et al., 2008; Stapleton et al., 2015; Stapleton et al., 2018). Participants rated the extent to which the intervention was interesting, understandable, useful, and positive with item response options ranging from 0 (not at all) to 10 (extremely).

Engagement.

Intervention engagement was measured in two ways. First, data on intervention completion rates and responses to intervention questions recorded by the intervention software provided program analytic indicators of engagement. Second, self-reported engagement was collected via an online survey following the completion of the intervention. Survey items from the Audience Engagement Scale (AES) (Greene et al., 2015) were used to measure key aspects of engagement with intervention content. Items in the original AES were designed to capture user perceptions to an in-person, curriculum-based intervention so we chose the most relevant items and adapted the wording to match the web-based intervention delivery format (e.g., “the workshop” was adapted to “the program”). The AES scale measures three constructs: active involvement (i.e., participants’ depth of engagement with the program), perceived novelty (i.e., participants’ perceptions of the newness or originality of the intervention), and personal reflection (i.e., the degree to which knowledge acquired is to re-evaluate personal conduct). All items (shown in Table 3) were measured using a 5-point Likert scale with response options ranging from 1 (strongly disagree) to 5 (strongly agree).

Table 3.

Mean scores for self-reported intervention engagement measures.

Construct	Survey Item	Meana (SD)
Active Involvement	The program said something important to me.	4.3 (0.9)
	I enjoyed this program.	4.6 (0.8)
	This program was boring (reverse coded).	3.9 (1.4)
Perceived Novelty	I’ve never done anything like what I did in the program.	4.3 (1.1)
	The information was different than information I have received about tanning in the past.	4.0 (1.1)
Personal Reflection	The information in the program made me think about my tanning.	4.5 (0.8)

Note.

^aSurvey item responses range from 1 (strongly disagree) to 5 (strongly agree).

Primary Outcomes.

The primary outcomes of indoor tanning bed use, sunburns, and indoor tanning intentions were assessed at both the baseline and follow-up surveys. Participants were asked to recall the number of indoor tanning sessions they had in the past two months. A similar question format was used to assess the number of sunburns they received in the past two months. Sunburns are a commonly-experienced side-effect of tanning bed use and independent risk factor for skin cancer risk (Stapleton et al., 2013). Our measure did not specify if reported sunburns were a result of indoor tanning so the measure is considered to be a general marker of ultraviolet risk behavior. For tanning intentions, participants indicated how likely they were to use an indoor tanning bed in the next year on a 6-point response scale anchored with 1 = Extremely unlikely and 6 = Extremely likely (Hillhouse et al., 2008).

Exploratory Outcomes.

Interest in changing tanning was assessed using a single item (Would you like to reduce or quit indoor tanning if you could do so easily?) adapted from Sobell and colleagues’ (1996) questions about readiness to change drinking behaviors (responses coded as 0 = no and 1 = yes). Self-efficacy in quitting tanning was assessed using Zeller and colleagues (2006) single item assessing perceived difficultly in quitting tanning (How hard would it be for you to stop using tanning beds/booths? (Please answer on a scale for 0–10, where 0 is “Not at all hard”, 10 is “Extremely hard”).

Data Analyses

Acceptability and engagement metrics are described for all participants who completed the intervention evaluation. We compared intervention and control participants on their baseline responses for the primary study outcomes and found evidence of non-equivalence. Specifically, intervention participants reported disproportionally higher rates of lifetime indoor tanning use (M _intervention = 409.0, M _control = 158.5) and 12-month indoor tanning (M _intervention = 72.3, M _control = 49.0) compared to control participants. A series of 2 (condition) X 2 (time) mixed-measures analysis of variance (ANOVA) was used to examine differences in study outcomes to control for baseline differences between conditions. We report partial eta square effect size estimates which can be compared to benchmarks defined by Cohen as small (𝜂_partial²= 0.01), medium (𝜂_partial² = 0.06), and large (𝜂_partial² = 0.14) (Laken, 2013). Our power calculations indicated that a sample size of 54 participants would allow us to detect a moderately large intervention effect (specifically, 80% power to detect an 8 indoor tanning session difference between conditions).

Data Sharing Statement

Study data were collected prior to the initiation of journal data sharing requirements. Participants were not asked to provide permission for their data to be publicly shared in the study consent form so data has not been shared with the article. However, data analysis syntax and output files were provided and the protocol and analysis plan were registered on clinicaltrials.gov prior to data collection.

RESULTS

Participants

A total of 6,264 individuals were assessed for eligibility to participate in the study, primarily through Qualtrics online screening. As detailed in the CONSORT flow diagram (Figure 1), more than 98% of those screened did not meet study inclusion criteria. Study email invitations were sent to the 91 eligible participants and 54 completed the baseline survey (intervention group n = 28, the waitlist control group n = 26). One participant was removed from the study after providing the same response for questions in the baseline survey (i.e., clicking strongly agree for all Likert items in the baseline survey). Participants’ (age M= 22.6, SD = 1.9) demographic characteristics are presented in Table 2.

Figure 1.

Consolidated Standards of Reporting Trials (CONSORT) Flow Diagram

Table 2.

Demographic characteristics of participants.

	Overall	By Assignment
	(n = 53)	Intervention (n = 27)	Control (n = 26)
Mean Age (SD)	22.6 (1.9)	22.9 (1.7)	22.3 (2.1)
Race
Caucasian/White	45 (84.9%)	24 (88.8%)	21 (80.8%)
Black or African American	3 (5.7%)	1 (3.7%)	2 (7.7%)
Other	5 (9.4%)	2 (7.4%)	3 (11.5%)
Ethnicity
Hispanic	8 (15.1%)	5 (18.5%)	3 (11.5%)
Non-Hispanic	45 (84.9%)	22 (81.5%)	23 (88.5%)
Education
High school graduate or lower	5 (9.4%)	1 (3.7%)	4 (15.4%)
At least some college	48 (90.6%)	26 (96.3%)	22 (84.6%)
Student Status
No, I am not enrolled in school.	21 (39.6%)	14 (51.9%)	7 (26.9%)
Yes	32 (60.4%)	13 (28.1%)	19 (73.1%)
Annual Household Income
$0 to $19,999	10 (18.9%)	6 (22.2%)	4 (15.4%)
$20,000 to $39,999	6 (11.3%)	3 (11.1%)	3 (11.5%)
$40,000 to $59,999	10 (18.9%)	5 (18.5%)	5 (19.2%)
$60,000 to $99,999	16 (30.1%)	8 (29.6%)	8 (30.8%)
Over $100,000	11 (20.8%)	5 (18.5%)	6 (23.1%)
Marital Status
Married/living with partner	18 (34.0%)	12 (44.4%)	6 (23.1%)
Single	35 (66.0%)	15 (55.5%)	20 (76.9%)

Acceptability

The means for the intervention evaluation items (measured on an 11-point scale with 0= not at all and 10= extremely) were: interesting 9.0 (SD = 1.6, range 1–10), understandable 9.3 (SD = 1.2, range 5–10), useful 8.7 (SD = 1.6, range 4–10), and positive 9.3 (SD = 1.1, range 6–10). Forty-three participants (90%) indicated they would recommend the website to a friend if it was publicly available.

Engagement: Program Analytic Metrics

The average intervention completion time was 17.2 minutes (SD= 11.7). There were a total of 83 unique questions programmed throughout the intervention that participants were asked to provide a response. Across all participants, 59 of these questions (71% of the total prompts) had no missing data and 19 (23%) had a missing response from only one participant. Each of the 4-week series of boosters administered to those in the intervention condition were completed by 80% of intervention participants (n = 20). Of the remaining five intervention participants, one viewed 3 of 4 boosters, one viewed 2 boosters, and three completed only 1 booster.

Engagement: Self Report

Strong engagement was reported on the post-intervention evaluation with high mean levels of agreement with all of the engagement questions including active involvement, perceived novelty, and reflection (Table 3).

Outcomes

A decline in 2-month tanning rates from baseline to follow-up was observed for both groups but the difference in the primary outcome of past 2-month indoor tanning at follow-up was not significant between study conditions (Table 4). Non-significant differences were also observed for indoor tanning intentions and sunburns. Perceived difficulty in quitting tanning was similar for control participants across time but a low mean was observed at follow-up compared to baseline for intervention participants. A moderately strong effect size was observed for the time X group interaction for perceived difficulty in quitting (η_p² = .07). A moderately strong effect was also observed for interest in quitting tanning which decreased across time for control but increased for from 57% at baseline to 65% at follow-up among intervention participants (η_p² = .05).

Table 4.

A comparison of intervention and waitlist control participants on trial outcomes.

	Marginal Means (SD)				Model Resultsa
	Control		Intervention		F value	P value	Effect sizeb
Outcome	Baseline	Follow-up	Baseline	Follow-up
Number of past 2-month indoor tanning sessions	16.2 (3.4)	7.9 (2.1)	15.8 (3.5)	11.5 (2.1)	0.60	.44	small (0.01)
Intentions to use indoor tanning	4.7 (0.3)	4.8 (0.3)	5.3 (0.3)	5.0 (0.3)	0.89	.35	small (0.02)
Number of past 2-month sunburns	1.7 (0.6)	1.1 (0.4)	1.4 (0.6)	1.9 (0.4)	1.87	.18	small (0.04)
Perceived difficulty in quitting tanning	4.9 (0.6)	5.0 (0.6)	6.3 (0.6)	5.0 (0.6)	3.27	.08	moderate (0.07)
Interest in changing tanning	0.63 (0.10)	0.46 (0.10)	0.57 (0.10)	0.65 (0.10)	2.58	.12	moderate (0.05)

Note.

^aTest statistic from the test of the interaction of time and condition in mixed ANOVA models.

^bPartial eta squared (η²).

DISCUSSION

The purpose of this proof-of-concept study was to evaluate participant acceptability, engagement, and preliminary outcomes of a PFI to promote tanning cessation among a sample of frequent indoor tanning bed users. Intervention acceptability ratings were highly favorable, indicating that participants found the intervention content to be interesting, understandable, useful, and positive. These favorable ratings are comparable to our prior tanning interventions that were delivered to young women with any recent history of indoor tanning, as compared with the frequent indoor tanners we focused on in the current study (Hillhouse et al., 2008; Stapleton et al., 2015; Stapleton et al., 2018). Program analytic measures of intervention engagement showed participants consistently provided responses to intervention prompts for information and the majority of participants completed all of the intervention boosters. Self-reported engagement was also encouraging. High average scores on active involvement and perceived novelty suggest that participants found the intervention to be interesting, relevant, and new. Engagement theory studies have demonstrated that education and intervention programs that can elicit such interest and perceived novelty are more likely to result in meaningful cognitive processing of intervention content among recipients (Greene et al., 2015). High average scores were also observed for personal reflection (i.e., the program “made me think about my tanning”), important evidence of the value of the intervention produced personal reflection on one’s behavior as a first step toward behavior change in the PFI framework. The favorable evaluations of the intervention and high engagement with the web content are encouraging signs that frequent indoor tanners may be receptive to receiving brief interventions designed to engage them to reflect on their tanning and consider behavior change. ​

There were non-significant differences in tanning rates at the follow-up between participants in the intervention and control conditions. Rates of tanning were lower at the follow-up for both conditions. One explanation for these reductions is that the majority of the follow-up surveys were completed in the summer months, when tanning rates are typically lowest. There was also little evidence that the intervention reduced intentions to use tanning beds or reduced sunburns which are commonly reported with indoor tanning bed use. There was evidence that the intervention may have led to reductions in perceived difficulty in quitting and increased interest in reducing or quitting tanning. These are promising findings because these changes are consistent with MI principles of impacting precursors to behavior change efforts and suggest that the intervention may have some value in getting the process of behavior change started. A more intensive or multi-component intervention may be required to change the indoor tanning behavior of frequent indoor tanners. For example, promising but short-term effects that have been demonstrated with intervention techniques such as facial morphing (Blashill et al., 2018) might be further sustained if the current intervention could be used to enhance buy-in and interest in tanning behavior change prior to delivery.

Our IT intervention is novel in our use of a PFI approach designed to motivate behavior change among individuals engaged in addictive behaviors. Strengths of this study include novel aspects of the intervention such as targeting frequent indoor tanners and utilizing an approach designed to build motivation to change. Our data is unique is demonstrating high intervention acceptability and engagement in high-risk tanners. Beyond the contribution to the skin cancer prevention literature, this study demonstrates the value of using survey software such as SNAP as a cost-efficient approach to generating a preliminary version of a website intervention for preliminary testing. The appearance of a website can be approximated by incorporating survey pages that include images, text, and branding rather than including only survey questions. These programs are particularly useful for pilot testing PFIs as basic skip patterns and data piping can be used to present personalized feedback pages based on users’ response to prior survey questions. This formative study has some notable limitations including a small sample size that was powered on the ability to observe effects of tanning rates but not the other preliminary outcomes. We experienced unanticipated delays due to over-estimating the positive screening rate of identifying frequent indoor tanners at the study onset, which led to a recruitment period that was extended by several months. These delays changed the timing of our follow-up period for most participants to fall within the summer months, which are typically months with low indoor tanning rates. It is also possible that eligible participants who signed up for the study were open to considering changing their indoor tanning behavior compared to those who declined, which could introduce bias into our sample. Finally, participants were paid to complete surveys including an intervention acceptability evaluation immediately following the website intervention which may have inflated engagement rates. Future research will explore uptake of the intervention when offered without incentives tied to evaluation.

In conclusion, this online, motivational tanning intervention targeted to frequent indoor tanners was acceptable and engaging to participants. Although there was little evidence of differences in the preliminary outcomes of indoor tanning, sunburn, and intention to tan among intervention participants relative to control participants, we report reductions in difficulty quitting tanning and increases in interest in changing tanning. These findings suggest that brief online PFIs alone may not be sufficient to produce behavior change among frequent indoor tanners but it may be valuable as a first step in the change process as part of a multi-component, tanning cessation intervention.