Abstract
Skin cancers are the most common malignancy in Australia. Regular sunscreen use can reduce the incidence of cutaneous squamous cell carcinomas and actinic keratoses and has been associated with reducing the incidence of basal cell carcinomas and melanomas. However, sunscreen effectiveness is limited by the failure of the population to use it routinely. Interventions that promote the daily application of sunscreen may reduce the morbidity, mortality, and economic burden associated with skin malignancies. We reviewed the literature that examines the effectiveness of interventions to increase routine sunscreen use and found that no one strategy has been shown to be clearly effective in adults and that relatively few studies have aimed to increase routine use in groups at extreme skin cancer risk. Future research should consider how interventions can be best designed and how sunscreen use is measured so that cost-effective, feasible strategies that result in improved sunscreen use in adults can be established.
Skin cancer is a significant public health concern worldwide (Lucas et al., 2006), but especially in Australia where skin cancers comprise the most frequently diagnosed malignancies and the highest cancer care costs (Australian Institute of Health and Welfare [AIHW], 2019; Fransen et al., 2012). More than half of Australians are diagnosed with skin cancer by the age of 70 years and Australia has the world’s highest incidence of melanoma (International Agency for Research on Cancer, 2018). Both ultraviolet (UV) A and B radiation are carcinogenic. In as little as 6 minutes of sun exposure (Damian et al., 2001), they can cause DNA damage (Agar et al., 2004) and suppress cutaneous antitumor immune responses (Damian & Halliday, 2002). Thus, protection against low-dose, incidental UV radiation exposure is important in protecting from the development of skin cancers on chronically exposed sites such as the face. Sunscreens have been shown to reduce the incidence of squamous cell carcinoma within 2 years and premalignant actinic keratoses within a few months. With long-term use sunscreens have also been associated with reduced incidence of basal cell carcinoma and melanoma (Green et al., 2011; van Der Pols et al., 2006).
The current sunscreen use recommendations in Australia and New Zealand are that sunscreen with a sun-protection factor (SPF) of 30 or more should be used when the UV index is forecast to reach at least three (Whiteman et al., 2019), alongside other forms of sun protection such as sun-protective clothing, broad-brimmed hats, sunglasses, and shade. The application should cover the face, ears, scalp, neck, and all parts of the body that are not covered by clothing (Whiteman et al., 2019). Sunscreen should be applied 20 minutes before sun exposure (Cancer Council Australia Keratinocyte Cancers Guideline Working Party, 2019) and reapplied two hourly or more frequently if activities involve water or sweating (Whiteman et al., 2019). These guidelines are similar to those outlined in many international policies (American Academy of Dermatology, 2013; British Association of Dermatologists, 2013). However, the Australia and New Zealand consensus statement differs from others by recommending that sunscreen application forms part of one’s daily routine to limit the effects of everyday sun exposure (Whiteman et al., 2019). Relevantly, these countries have a higher incidence of months with daily maximum UV index averages that equal or exceed three.
Despite Australia’s high levels of UV radiation (AIHW, 2019) and several public awareness campaigns, sunscreen use remains poor for Australian adults, even among those with multiple previous skin cancers (Chen et al., 2015). Regular and frequent promotion of sunscreen use at the individual level is necessary and likely to be cost-effective (Hirst et al., 2012), especially for high-risk patients. To do this, evidence-based interventions should be employed.
Recent systematic reviews examine sun-protection measures in general, often with a focus on specific population subgroups and settings. Few focus specifically on sunscreen use. Of those reviews which do address sunscreen use in adults, the studies included are limited in number and do not address varied interventions to increase sunscreen use across broad adult populations and settings. One review of behavioral counselling interventions linked to primary care identified seven trials assessing sunscreen use in adults of which four demonstrated an increase in this outcome (U.S. Preventive Services Task Force, 2018). Others narrowed their focus to text-messaging interventions, high melanoma-risk populations, or tourist settings (Chambergo-Michilot et al., 2020; A. Rodrigues et al., 2013; Wu et al., 2016). The interventions identified in these reviews were often education based or prompts.
Education is a valuable tool in changing behavior as knowledge is a crucial component of making informed healthy decisions. However, health behaviors are influenced by a range of other social and cultural moderators (Glanz et al., 2008; Kelly & Barker, 2016). Accordingly, the success of educational interventions is reliant on the presence of gaps in participants’ knowledge.
Interventions that utilize prompts aim to minimize forgetfulness, a frequently reported barrier to sunscreen use (Boggild & From, 2003; Diehl et al., 2020; Garbutcheon-Singh et al., 2016), and provoke sun-protective behaviors through messages that directly or indirectly remind participants to apply sunscreen.
Previous reviews note that interventions to increase sun-protection use are mixed in their efficacy, studies are heterogeneous and the existing evidence provided is scant and often of lower quality.
There have been numerous trials examining interventions to increase sunscreen use in adults. Given the heterogeneous nature of these trials and the complexity of interventions and outcome measures, a narrative review has been conducted to inform future research. This review explores some of the features of these studies including population selection, settings, measurement of outcomes, the behavior change techniques adopted, and some of the practical issues related to the design and success of these interventions.
Method
Papers selected for this review were identified following a PubMed search in May 2020 using keywords and phrases including sunscreen, sun protection, sun safety, use, compliance, adherence, intervention, and trial. Searches were limited to published English language papers in peer-reviewed journals available via open access or through the University of Sydney collection. Included papers described experimental studies that quantified the effect of an intervention on actual sunscreen use in adults. Papers that only presented composite sun-protection outcome measures that included other forms of sun protection or studies that did not measure actual sunscreen use, such as those that only measured sunscreen use intentions, were not included.
Each paper was reviewed for study features, participant characteristics and intervention components. The data extracted from each included study were the design, theoretical basis, follow-up period, outcome measures, country, setting, and season. Participant characteristics that were collected were sample size, demographics, and sunscreen use results. The interventions and control conditions described in each paper were analyzed and categorized using the behavior change technique (bct) taxonomy (Michie et al., 2013) and the theory and techniques tool (Human Behaviour Change Project, 2018). Trends in the extracted data were identified, with examples to be presented and discussed in this review.
Results
A total of 23 studies fulfilled the inclusion criteria. Supplemental Table 1 describes each study’s intervention components, theory, population, setting, follow-up period, and outcome measure descriptions and whether sunscreen use results were significantly affected by each study intervention. Almost all studies were randomized-controlled trials (n = 21). Fourteen papers described follow-up periods of 2 months or more, which is described as the average required number of days for habit formation for performing a behavior in the same context daily (Lally et al., 2010). With regard to the outcome measures of the included studies, 19 papers used only self-report measures. The remaining papers used a nonself-report measure of sunscreen use. Indeed, 14 studies described mostly female participants, while the four studies conducted in outdoor occupational settings described mostly male cohorts. However, 11 papers described a significant intervention effect on sunscreen use (Supplemental Table 2).
Trends
Education
Many studies (n = 18) utilized knowledge-based techniques such as shaping knowledge (n = 16) and natural consequences (n = 15), often in addition to other techniques for multicomponent interventions. Most studies that demonstrated a significant impact on routine sunscreen use utilized at least one of these techniques (n = 9). These techniques were also popular as control conditions. Knowledge-based interventions typically aim to increase participants’ understanding of how and when to use sunscreen and the potential consequences of unprotected UV exposure, namely skin cancer and UV-related aging, where a lack of understanding poses a barrier to sunscreen use. Knowledge-based techniques were delivered via multimedia (n = 7), written (n = 3), verbal modes (n = 2), or a combination (n = 6).
Prompts
Several trials used prompts (n = 12), which fall under the associations BCT cluster (Michie et al., 2013), as stimuli to cue sunscreen use (Human Behaviour Change Project, 2018). Armstrong et al. (2009), Dixon et al. (2007), Duffy et al. (2018), and Gold et al. (2011) all designed intervention conditions that allowed for the isolation of the effect of electronic prompts on routine sunscreen use.
Using daily, fixed-frequency text messages containing weather forecasts and sunscreen use reminders, Armstrong et al. (2009) demonstrated a significant effect (p < .001) on daily sunscreen application (Supplemental Table 2), albeit after only a 6-week study period. At 6 weeks, daily sunscreen use was 56.1%, and 30.0% in the intervention and control groups, respectively, as measured by electronic monitoring. Messages were not personalized, targeted, or tailored to participants. While it is not known if these effects persisted beyond the study period, tolerability of the intervention appeared to be high with 69% of participants reporting that they would like to continue to receive the reminders.
Dixon et al. (2007) used weekly weekend weather forecast emails with UV index and behavior prompts, Duffy et al. (2018) used positively framed sun-protection text messages on three random weekdays per week, and Gold et al. (2011) used fortnightly humorous sun-protection text messages on varied days; however, all failed to demonstrate a significant impact on their respective measures of sunscreen use.
Environmental Context and Resources
Some interventions (n = 5) altered the environment surrounding or resources available to participants to facilitate a certain behavior or minimize barriers to performing the behavior. Researchers may utilize BCTs from the antecedents cluster such as adding objects to or restructuring the physical environment when conducting studies in fixed settings such as workplaces or outdoor recreational venues. When addressing routine sunscreen use, such interventions often increase access to sunscreen.
Duffy et al. (2018) isolated the effect of the provision of sunscreen on participants. While their text message condition alone did not demonstrate an increase in sunscreen use, the mailing of sunscreen to participants, when combined with thrice-weekly text messages, led to a mean difference in self-reported sunscreen use of 0.8, on a 5-point scale, which was significantly greater than both education-only, and education and text message groups.
All of Glanz et al. (2002), Huang et al. (2013), and Mayer et al. (2007) increased participants’ access to sunscreen as part of multicomponent interventions. All demonstrated significant impacts (Supplemental Table 2) on the self-reported sunscreen use of their respective populations.
Wang et al. (2017) took a novel approach to encourage sunscreen application. They hypothesized that changing the accessibility of sunscreen in the home would increase daily compliance. Their intervention was a container for storing toothpaste and sunscreen that altered the location and accessibility of sunscreen. While the study was small (n = 62) there was a nonsignificant trend in mass of sunscreen used (p = .06), with the intervention group using 20% more sunscreen than the control group over 6 weeks.
Health Behavior Theories
Sunscreen behavior change interventions tend to be founded in one or more psychosocial theories of health behavior change, such as social cognitive theory or social learning theory (n = 6), the health belief model (n = 4), the transtheoretical model (n = 3), Diffusion of Innovations (n = 2), the precaution adoption process model (n = 2), the theory of planned behavior (n = 2), and health action process approach (n = 2). Outcomes vary between studies but do not appear to be related to the use of one theory or another as studies with a significant intervention effect on sunscreen use have a similarly diverse suite of models employed to those that did not demonstrate such results.
High-risk Populations
Of the studies included, five targeted high-risk groups such as melanoma patients and their relatives (Bowen et al., 2015; Bowen et al., 2019; Geller et al., 2006) or individuals identified using a risk assessment tool (Glanz et al., 2010; Glanz et al., 2015). All five studies tailored their interventions to individual participants’ characteristics, such as risk factors and behaviors; however, only the SCAPE (Glanz et al., 2010) and PennSCAPE (Glanz et al., 2015) trials demonstrated significant effects (Supplemental Table 2) on self-reported sunscreen use, represented through both diary and mean frequency of sunscreen use among participants as measured by ordinal scale.
Outcome Measures
Self-report of sunscreen use, typically using ordinal response, was the most popular form of outcome measure (n = 22). For a number of these self-report measures, it was unclear as to whether the tools used were validated. Moreover, the wording of these tools varied between studies (Supplemental Table 1) in the description of sunscreen use and the available responses for participants to provide. Many measured the frequency of sunscreen use, however, various provisions were applied in different studies. These included details of the SPF used, the circumstances in which sunscreen was applied, and the period over which the participants should recall their sunscreen practices. Of the eight papers that described the required SPF of sunscreen to constitute targeted outcomes, all set the threshold at SPF15. Reapplication was rarely assessed.
Similarly, how authors ordinally defined sunscreen use as an outcome varied. Where some presented descriptive data of the number of participants for each response, authors that dichotomized their results did so by comparing often and always or always only with all other responses.
Objective measures of sunscreen use were used by four studies, three of which were used in conjunction with a self-report measure. Wang et al. (2017) used the difference in sunscreen weight before and after their intervention as a measure of compliance. Armstrong et al. (2009) used electronic monitors to report the times of sunscreen vessel cap opening and closing.
Another objective measure was utilized by Mayer et al. (2007) and Szabó et al. (2015) to strengthen the validity of their self-report data. Mayer et al. (2007) used a colorimeter device to measure skin color of participants’ faces on spectra of black–white, and blue–yellow. They demonstrated a significant (p = .009) group-by-time interaction in blue–yellow measurements of the participants’ faces, suggesting that control group participants were more tanned than those in the intervention group over 2 years. Szabó et al. (2015) used a Mexameter to measure the level of melanin and erythema on the volar forearm of participants as an indicator of sunscreen use. They found no significant differences in change of erythema or melanin levels between groups, contradicting the significant difference (p < .05) in self-reported sunscreen use between groups (Supplemental Table 2).
Discussion
Education
This review suggests that educational interventions may be useful for improving the sunscreen use of specific populations, such as those at high risk for skin cancer. These high-risk groups may benefit from tailored interventions that address participant characteristics to heighten the salience of the consequences of sun exposure. This is in keeping with a previous review that suggested increasing the salience of risk may prompt preventive behaviors in adults with moderate risk for melanoma (Wu et al., 2016).
Prompts
Electronic reminders were a common medium for interventions that utilized prompts to increase routine sunscreen use. Text-message interventions have been used for a range of preventive health behaviors and are moderately efficacious (Head et al., 2013). While these interventions can be low cost and easy to disseminate (Dixon et al., 2007; Gold et al., 2011), authors cited technical difficulties (Gold et al., 2011), and queried the utility for populations with irregular device access (Dixon et al., 2007). Similarly, poor understanding or concerns about adverse effects, systemic absorption (Matta et al., 2019), and contact reactions (Heurung et al., 2014) are barriers to sunscreen use (Diehl et al., 2020) that may not be overcome by simple prompts (Armstrong et al., 2009). While forgetfulness is a common barrier, other barriers exist that may limit the efficacy of prompt-focused interventions (Armstrong et al., 2009). Furthermore, correct sunscreen use requires reapplication at least two-hourly while in the sun, which may be harder to prompt due to the variability in schedules of participants.
Gold et al. (2011) found that fortnightly text messages were sometimes perceived by a young–adult population as being uninteresting or annoying. While other studies did not demonstrate this sentiment exactly, the opt-out system in some of these interventions saw large numbers of participants withdraw from receiving the intervention (Duffy et al., 2018; Gold et al., 2011).
Reviews of text messaging for preventive health behaviors have identified several intervention design features that are moderators of their efficacy. These moderators include personalization, such as using the participant’s name, targeting messages to characteristics of population subgroups, tailoring messages to participant demographic or psychosocial variables, decreasing or user-determined message frequency and a 6- to 12-month intervention duration (Armanasco et al., 2017; Head et al., 2013). As such, the design of such interventions should consider these features. As most of the prompt-based studies were conducted before the aforementioned reviews were published, few of the interventions were designed with the optimal use of these moderators in mind and thus they were largely overlooked. None of these interventions were targeted or tailored, and only the emails used by Dixon et al. (2007) were personalized. Similarly, all intervention periods used in these studies were shorter than the recommended 6 to 12 months.
Environmental Context and Resources
All of Glanz et al. (2002), Huang et al. (2013), and Mayer et al. (2007) increased participants’ access to sunscreen as part of broader interventions and all demonstrated significant impacts on their participants’ self-reported sunscreen use. This is noteworthy given that barriers to sunscreen access, such as cost, are less frequently cited than more established barriers to use such as forgetfulness or discomfort (Boggild & From, 2003; Diehl et al., 2020).
The sunscreen–toothpaste storage unit approach taken by Wang et al. (2017) allowed them to alter the home environment of participants which can both increase the situational accessibility of sunscreen and increase awareness of the need to apply sunscreen and prompt action. An intervention like this could be inexpensive and, assuming continued use, participants could continue to benefit from the storage unit after the trial period which may aid in facilitating habit formation.
High-risk Populations
Trials aimed at high-risk populations make up a minority of published studies of interventions to increase sunscreen use. Such interventions are valuable as they allow for a cost-effective allocation of resources for skin cancer prevention (Glanz et al., 2003). Furthermore, previous studies have found that sunscreen use is low for melanoma survivors and their families (Bowen et al., 2012; Manne et al., 2004; V. M. Rodríguez et al., 2017) demonstrating the importance of interventions that target these groups. It should be noted that the risk factors addressed for these trials favored risk of melanoma and not keratinocyte cancer (Bowen et al., 2019; Geller et al., 2006; Glanz et al., 2010; Glanz et al., 2015). Moreover, publications aimed at increasing sunscreen use in patients with extreme keratinocyte cancer risk, such as organ transplant recipients (Leung et al., 2018), are few.
Women tend to make up the majority of participants in sunscreen compliance trials, threatening the generalizability of findings when these interventions are employed for men, who have higher risks of and mortality from skin cancer (AIHW, 2019; Kasparian et al., 2009) and may be less likely to use sunscreen (Diehl et al., 2020).
Outcome Measures
As stated by guidelines, sunscreen represents one of several sun-protection measures that people should utilize (Whiteman et al., 2019). This review demonstrates that what separates sunscreen from these other measures in research is the inconsistency in the definition of sunscreen use as well as the limited ability to capture reapplication and the quality of sunscreen application. Moreover, the use of SPF15 as the threshold for acceptable SPF is not in keeping with current recommendations (Whiteman et al., 2019).
Skin malignancies on the head and neck have a greater risk for metastasis or recurrence and treatment is more likely to have cosmetic impacts (Cancer Council Australia Keratinocyte Cancers Guideline Working Party, 2019). While sunscreen may be applied more often to these sites (Gorig et al., 2020), the application is often heterogeneous with areas around the hairline and ears commonly left exposed (Lademann et al., 2004; Pratt et al., 2017). As such, it might be reasonable to prioritize and provide instructions for the correct application of sunscreen on these sites. Despite this, detailed descriptions of intended sunscreen application by participants are rarely featured in these papers.
Sunscreen use outcomes are commonly measured via self-report (see Supplemental Table 1). This is often performed with ordinal scales, such as those where scores of one to four are allocated to responses never through to always, or diarized sunscreen use where participants record their use over a prescribed period. Ordinal representations of sunscreen use are difficult to interpret in reference to sun-protection guidelines which are specific and quantitative. Moreover, self-report measures of sunscreen use can be imprecise and are known to overestimate adherence (Haynes et al., 2002).
An example of the shortcomings of self-report measures of sunscreen use was demonstrated by Glanz et al. (2010) who found no difference in sunscreen use between intervention and control groups on a 4-point ordinal scale, but a significant difference in adjusted mean change scores of diarized sunscreen use between groups (see Supplemental Table 2). Research into the validity of self-reported sunscreen use found that diary assessment was better correlated with skin-swabbing, an objective measure of sunscreen use, than the scale method (Glanz et al., 2009). Despite this, ordinal scale surveys remain the prevailing measure of sunscreen use.
Armstrong et al. (2009) used an ordinal scale measure alongside electronic monitors and found that both the intervention and particularly control group overestimated their sunscreen adherence. However, while sunscreen mass and electronic bottle monitors may be effective sunscreen use measures, with electronic monitors considered the gold standard for measuring treatment adherence (Feldman et al., 2017), both can be faulted. Electronic monitors can be subject to the opening and closing of the cap without any sunscreen dispensed (Armstrong et al., 2009), and there is no assurance that a change in mass of sunscreen equates to sunscreen use on the prescribed area of the participant.
Skin color quantifiers were another objective measure used in the included studies. However, these tools are only used on specific areas of the skin of participants (Mayer et al., 2007; Szabó et al., 2015), thus they can only truly measure UV exposure to this area, and only if UV exposure exceeds the measurable threshold for sunburn and/or pigmentation in participants whose skin can tan. Sunscreen use could also influence and increase the duration of sun exposure (Autier et al., 2007) which may complicate the interpretation of these devices’ readings.
Limitations
This review has several limitations. First, the literature search was not systematic nor exhaustive, and only articles in PubMed were included. The inclusion of only peer-reviewed literature in English may have excluded some influential studies, while the lack of a systematic evaluation of the quality of each study may mean that interpretation and comparison of the findings of each study may not be weighted appropriately. Similarly, the inclusion of only published papers can be complicated by publication bias. Another limitation specific to the review of sunscreen use trials was that several papers were eliminated where they only presented composite sun-protection behavior outcomes. This may have facilitated the elimination of studies that examined interventions that effectively increased routine sunscreen use.
As routine sunscreen use is this article’s main outcome of concern, other outcomes such as the efficacy of sunscreen to prevent skin cancer, as described in detail elsewhere (Rueegg et al., 2019; saes da Silva et al., 2018) were not examined. As a result, an investigation of the effects of interventions on skin cancer incidence falls outside of the scope of this review.
Conclusion
Trials of interventions to increase sunscreen use to date are mixed in their approach and implementation. These interventions adopt a range of and often multiple behavior change techniques. Promising interventions include reminders and personally tailored education and those that alter participants’ environmental context or available resources; however, no single approach has proven to be effective across a range of patient populations. This review identifies that there are grounds for further research to address long-term sunscreen use and reapplication, especially in high skin cancer risk populations and men. While many trials use subjective measures of sunscreen use, these are difficult to interpret for clinical application and comparisons of results between studies are limited by their varied definitions of sunscreen use. Future research should seek to utilize objective measures of sunscreen use for clear interpretation of results.
Supplemental Material
sj-docx-1-heb-10.1177_10901981211046534 – Supplemental material for Interventions to Increase Sunscreen Use in Adults: A Review of the Literature
Supplemental material, sj-docx-1-heb-10.1177_10901981211046534 for Interventions to Increase Sunscreen Use in Adults: A Review of the Literature by Nicholas Allen and Diona L. Damian in Health Education & Behavior
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
