The study was conducted as an online baseline survey of a randomized controlled trial—the mySmartSkin web-based intervention designed to increase skin self-examination and sun protection behaviors among melanoma survivors. A detailed description of the mySmartSkin study, including recruitment, measures, and other analyses, is available elsewhere (Coups et al., 2019); Manne et al., 2020). The Institutional Review Boards affiliated with the University and Medical Center approved this study.
Participants were adult melanoma survivors who were recruited through four different sites within one US state: a National Cancer Institute-designated comprehensive cancer center; a Department of Dermatology; a Medical Center; and a State Cancer Registry. Potentially eligible participants received a study information letter and consent form, and a member of the research team attempted to contact each patient to determine their eligibility. For patients recruited through the NJSCR, an information letter was mailed to each patient’s physician requesting that the physician contact the research team within two weeks if there were any reasons that the patient should not be contacted for the study. The research team at the Cancer Institute determined the eligibility of patients who expressed interest to take part in the study. Of the 1,411 individuals assessed for eligibility and study interest, 150 (10.6%) were deemed ineligible, and 776 (55%) declined to participate, with the most common reason provided being lack of interest (n = 399). Of 485 patients who consented to the study, 441 (90.9%) participants completed the online survey. Upon completion of the baseline survey, each participant received a $25 gift card.
Inclusion criteria to participate in the study were: a) diagnosed with primary pathologic stage 0-III cutaneous malignant melanoma; b) were 3 to 24 months post-surgical treatment; c) were not performing thorough SSE (i.e., they did not report inspecting each of 15 areas of the body at least once during the past two months) (Weinstock et al., 2004) and/or not following sun protection recommendations (i.e., a mean score of < 4.0 on a sun protection behavior index that assessed the frequency of engaging in four behaviors, each assessed on a 5-point scale from 1=never to 5=always (Glanz et al., 2008); d) were 18 years old and above; e) had Internet access; f) able to read and speak in English and g) able to provide informed consent.
Outcomes. Four individual sun protection behaviors (use of sunscreen with sun protection factor [SPF] 30 or more, wearing a long-sleeved shirt, wearing a wide-brimmed hat, and staying in the shade or under an umbrella) were assessed over the past 12 months (e.g., “In the last 12 months, when you were outside on a sunny day, how often did you wear a long-sleeved shirt?”) (1 = never to 5 = always) (Glanz et al., 2008). The mean of these four items was calculated to create a total index of sun protection behaviors in the past 12 months.
Demographics. Participants reported their age, education level, marital status, and sex.
Medical factors. Medical records were used to extract months since surgery and disease stage at diagnosis (stage 0, stage 1, stage 2, and stage 3).
Skin cancer risk factors. Participants reported the presence or absence of eight skin cancer risk factors, which resulted in a total score from 0 to 8: light eye color, light natural hair color, fair untanned skin, skin sensitivity, presence of freckles, presence of large moles, ever indoor tanning, and family history of melanoma (Bränström et al., 2010; Manne & Lessin, 2006). Participants also reported how many times they had experienced a “red, pink, or painful sunburn that lasted a day or more” in the last year, which was dichotomized into zero versus one or more sunburns.
Melanoma knowledge and attitudes.Melanoma knowledge was assessed through 13 items (e.g., “Melanoma is the most common form of skin cancer”) (Coups et al., 2016; Gillen et al., 2011; Manne & Lessin, 2006). The number of correct responses created the total score, Cronbach’s alpha (a) = .693. Attitudes about melanoma included perceived severity (6 items, a = 0.871), perceived controllability (4 items, α = 0.669), and perceived risk for melanoma recurrence (4 items, a = 0.814) (Manne & Lessin, 2006; Vickberg, 2003).
Psychological measures.Distress about melanoma was determined by asking participants to select a number that best describe how distressed they are currently about their melanoma (1 = not at all distressed to 10 = extremely distressed) (Manne & Lessin, 2006). Worry about melanoma recurrence was assessed through the mean of four items, α = 0.918 (Moss-Morris et al., 2002).
Social influence. Social influence factors included physician recommendation about sun protection behaviors (4 items, a = 0.741), descriptive norms regarding sun protection behaviors, and injunctive norms regarding sun protection behaviors (Coups et al., 2019; Manne & Lessin, 2006). Descriptive norms about sun protection behaviors were determined through the mean of five items (e.g., “My friends and family use sunscreen with a SPF of 30 or more when they are outside on a sunny day”) (5 items, a = 0.772). Injunctive norms about sun protection behaviors were assessed by the mean of five items (e.g., “My friends and family think I should use sunscreen with a SPF of 30 or more when they are outside on a sunny day”) (5 items, a = 0.861).
Attitudes toward sun protection behaviors. These were measured through perceived benefits regarding sun protection behaviors (12 items, a = 0.902) perceived barriers regarding sun protection behaviors (23 items, a = 0.903), and self-efficacy underlying sun protection behaviors (12 items, a = 0.938) (Azzarello & Jacobsen, 2007; Bränström et al., 2010; Coups et al., 2011).
Data were analyzed using IBM SPSS Statistics (version 25). Univariate analyses were conducted to assess the relationship between the demographic variables and sun protection behaviors. Hierarchical multiple regression analyses were conducted to evaluate predictors of sun protection behaviors. Variables were entered into the model in seven steps: demographic variables were entered first, then cancer-related factors, skin cancer risk factors, knowledge and attitudes, psychological characteristics, social influence, and attitudes toward sun protection behaviors (perceived benefits, perceived barriers, self-efficacy). Change in R2 was computed at each step. Statistical significance was determined at an alpha level of 0.05.