Participants and procedures
Ethical approval of the study protocol was granted by the Institutional Review Board of the Faculty of Medicine, McGill University (reference no. A11-B39-11B). Eligible patients will include English- and French-speaking adult patients with a confirmed diagnosis of melanoma who seek services at McGill teaching hospitals. Together, these hospitals treat over 400 new melanoma patients annually [77, 78]. Previous psychosocial studies with cancer populations at the same hospitals [30, 79, 80] and at other sites [81–92] reported participation rates between 57-86%, with a mean of 76%. Attrition rates range between 4% and 24% in psychosocial studies with an assessment up to 18 months after the treatment of patients with melanoma [56, 93] or other cancers [67, 82, 83, 94–97]. Enrollment will take place over 2 years and continue until 200 participants have completed the study. A sample of 200 participants provides sufficient statistical power for the analyses (for power estimates see section Data analysis). At all time points the questionnaires will be provided as paper-pencil versions. Participants will be asked up to 3 times to mail back questionnaires completed at home. Interviews will be conducted over the telephone.
During regular clinic visits up to 12 months after their melanoma diagnosis, i.e., at time 1 (T1), eligible patients will be advised by the clinical care staff about the opportunity to participate in the study. In addition, the study will be advertised through flyers and posters in the waiting room area, which will allow interested patients to proactively contact the research team about study participation. The research assistant will provide study information to interested patients, verify study eligibility and collect written, informed consent. Participants will be asked to complete baseline questionnaires on sociodemographic and illness-related information, past SSE behaviors, and psychological functioning at the clinic. Alternatively participants can complete the questionnaires at home and return them to the research team by mail. The research assistant (RA) will access the participants’ medical charts in order to complete the medical information sheet. The second assessment will take place in conjunction with the delivery of the dermatological education at the clinic 3 to 6 months after T1, i.e., at time 2 (T2). Participants will complete questionnaires about distress, coping strategies, physician support and about SSE knowledge, attitude, and self-efficacy. In addition, participants who, at the time of data collection, report having a spouse (i.e., having a committed intimate relationship), will be asked to answer questions related to their partner’s impact on SSE practice. Furthermore, these patients will be asked to take home a survey for their spouse, which mirrors the partner-related questions that the patient is asked to answer regarding SSE. Also at time 2, all study participants will receive a 20-minute standardized dermatological education on SSE derived from empirical evidence [14, 29, 47, 50, 65, 98–104] and best practice guidelines [36, 38, 105–109]. Patients will be encouraged to attend the education session together with a significant other (e.g., spouse, other family member, friend) who could assist them with SSE. During the education session, the dermatology-trained RA, i.e., the health educator, will emphasize the usefulness of monthly whole body exams as an effective measure to detect suspicious skin lesions as early as possible. The educator will provide detailed information about how to conduct an effective SSE. This will include an explanation of the well-established ABCDE paradigm (lesion Asymmetry; Border irregularity; Color variation; Diameter; Evolution, e.g. change in size, shape, symptoms, etc.) [110–113] for the detection and interpretation of pigmented lesions by lay persons and health care professionals. Furthermore, the educator will provide handouts, which will assist participants with regular SSE, including a summarizing brochure on melanoma , a bookmark with color printed examples of lesions , a leaflet providing the link to an online video modeling skin self-examination , and a SSE Journal to record skin spots of concern and body parts covered during each home SSE [116, 117]. At the end of this 20-minute session, the educator will encourage the participant to reflect on their SSE intentions. Subsequently the participant will be asked to note down in the SSE journal if, when, where, and assisted by whom the participant plans to conduct self-exams. At time 3 (T3), i.e., 3 months after the dermatological education, participants are invited by letter to fill out questionnaires assessing distress, coping strategies, physician and spousal support as well as SSE knowledge, attitude, and self-efficacy. In addition, a 10-minute, structured SSE behavior interview with the RA will be scheduled for T3 to assess practice of SSE over the last 3 months. At time 4 (T4), 12 months after the dermatological education, participants will be invited by letter to complete an assessment identical to T3, i.e., questionnaires and SSE behavior interview. At T4 a subsample of 30 patients, which will include the first 15 men and the first 15 women who consent to participate, will be invited to take part in a 50-minute individual, semi-structured interview based on the McGill Illness Narrative Interview  focusing on the following questions: 1) how patients deal with the diagnosis and how they experience the illness; 2) their experience regarding SSE, including how they deal with the ongoing need for thorough SSEs and obstacles to SSE; 3) their experience of physician support of SSE, or lack thereof, and how they deal with it; and 4) their experience of partner support of SSE, or lack thereof, and what facilitates or hinders partner support and how they deal with it. Biological sex is considered explicitly given that previous melanoma research suggests that this is a key variable in relation to SSE [27, 45, 119–121]. Participants will be invited to share and report their thoughts, feelings, and behaviors in each of the 4 domains, which were selected based on their relevance to the cancer trajectory and their presumed significance for SSE practice.
Descriptive and independent variables: predictors of skin self-examination
Study measures were selected based on their wide use in cancer research, on their psychometric properties, and on their direct relevance to the objectives of the project. They are used to assess descriptive and independent variables, which are expected to have an effect on SSE behaviors during the first 3 months (T1) and 12 months (T2) after dermatological education.
The Sociodemographic Information Form  includes questions about age, sex, education, having a spouse, socioeconomic status, and cancer diagnosis. With the Medical Information Sheet  the RA gathers data such as time since diagnosis, melanoma stage and depth, previous diagnosis of cancer, melanoma treatment and disease progression. The Skin Cancer Prevention Scale  captures whether patients performed SSE and have been advised by health care providers about SSE prior to their current melanoma diagnosis. The Skin Cancer Knowledge Scale, based on questionnaires by Hay et al. 2006  and Manne et al. 2006 , assesses knowledge regarding melanoma risk, melanoma warning signs and SSE. The SSE Attitude Scale, an adaptation of Manne’s SSE Benefits and Barriers Scale , assesses perceptions of SSE importance, personal gain through SSE, and barriers to SSE. The SSE Self-Efficacy Scale, based on Weinstock et al. 2007 , captures an individual’s self-confidence in performing effective skin self-exams. The Physician SSE Support Scale  inquires about the perceived interest that the treating physician conveys regarding the patient’s practice of SSE. The Berlin Social Support Scale  assesses the patient’s perception of emotional, instrumental and informational illness support provided by their spouse (patient perception and partner perception). The Skin Cancer Index [126, 127] is a self-report scale focusing on emotional, social and appearance-related concerns associated with skin cancer. The Patient Health Questionnaire, PHQ-4,  screens for symptoms of depression and anxiety. The COPE Inventory  inquires about an individual’s use of coping strategies, e.g., mental disengagement, behavioral disengagement, focus on and venting of emotions, use of instrumental social support, denial, humor, use of emotional social support, planning and others.
Dependent variables: skin self-examination
Adherence to medical advice on SSE is assessed at 3 months (T3) and at 12 months (T4) after the standardized education session. Based on prior SSE research [30, 45, 58] and on recommendations for a monthly, comprehensive skin self-exam from the Canadian Dermatology Association , the American Academy of Dermatology  and the American Cancer Society , a structured SSE-behavior interview was developed and pilot tested . During this 10-minute interview the RA records SSE behaviors in terms of SSE completeness and frequency. With the help of a calendar and the body map, which participants receive during the education session in order to document each SSE at home, SSE is recorded for five specified areas of the body: 1) head and neck; 2) front upper body including arms and shoulders; 3) front lower body including groin/genital area, legs, and feet; 4) back upper body including lower back; 5) back lower body including buttocks and back of legs. SSE assistance by significant others and the use of melanoma pictures during the self-exams will be documented. The first dependent variable, Completeness of SSE, refers to the examination of the 5 body areas mentioned above. The RA inquires about each of these 5 parts of the body for each skin self-exam reported by the patient over the last three months at T3 and at T4. The interviewer gives one point for each body area examined, for a possible total of 5 points reflecting the completeness, in regards to the 5 body areas, of each SSE conducted. A mean score (on 5) is calculated across all SSEs. The second dependent variable is Frequency of SSE over a given time period. Over the period of 3 months covered by the assessment at T3 and at T4, ideal frequency would be 3 SSEs, with more than 3 SSEs reflecting too high a frequency, and less than 3 reflecting too low a frequency. The total count of SSEs (i.e., how many times a patient performed a SSE) over 3 months will be used to determine if the frequency with which an individual patient conducted SSEs was too low, ideal or too high, which will be used for the data analysis.
In all relevant cases, additional analyses, including interaction effects and subsample differences, will be examined. Data analyses related to objective 1 ( i.e., determining the extent of thorough SSE performance defined in terms of frequency and completeness): Descriptive and inferential statistics will be employed to gain a differential picture of patients’ practice of SSE in terms of completeness and frequency at T3 and T4 for the total sample as well as for subsamples according to sex, age, having a partner, etc. Data analyses related to objective 2 (i.e., identifying psychosocial variables independently associated with thorough SSE): First, hierarchical multiple regressions will be performed with SSE completeness as dependent variable using T3 data, i.e., 3 months after standardized dermatological education. The independent variables described above will be entered block-wise, i.e., 1st block: sex; 2nd: skin cancer knowledge, attitude toward SEE; 3rd: self-efficacy for performing SSE; 4th: skin cancer-specific distress measure; 5th: partner support and physician support; 6th: total score of the coping measure (note that analyses will be repeated with the individual coping subscales). Second, a multinomial regression will be conducted with the same predictor variables and SSE frequency (i.e., too low, ideal and too high frequency) as the dependent variable. The strategies described for T3 will be repeated using the data collected at T4, i.e., 12 months after the education session. With the projected sample size of N = 200, a minimal R2 increment of .08 will be detected with .79 power. The power for detecting a R2 increment of .09 will be .85. The variance explained by both blocks and individual variables will be examined. The two dependent variables are treated separately because they reflect different aspects of SSE practice. As a second step, they may also be clustered and the resulting composite measure of SSE behavior will serve as the dependent variable in regressions similar to the ones described above. Data analyses related to objective 3 ( using qualitative methods for an in-depth understanding of psychosocial factors related to SSE): Data saturation in qualitative theme analysis, depending on the level of structure of the interview, is often achieved with a sample of 10–15 participants who are examined intensively, while larger samples typically add only minimal new data . Consensual Qualitative Research [131–133] will be applied to the semi-structured interviews of 15 male and 15 female patients. This qualitative method is based on principles of grounded theory  and aims at developing a theory directly grounded in the phenomena under study. As a first step, the 2 independent sets of interviews will be theme analyzed to establish preliminary categorizations through open coding. To guide this, the following preliminary questions will be used: how do patients cognitively, affectively and behaviorally adjust to cancer diagnosis and treatment?; how and why do patients engage in or refrain from performing SSE?; how do they experience practicing SSE (e.g., does SSE evoke discomfort, does it trigger tumor fear, does it lead to a sense of safety and control?); in which ways do significant others support or undermine SSE practice?; do patients find it hard to ask significant others for help with SSE or do patients feel overwhelmed by a others’ motivation to help?; and in which ways are physicians experienced as helpful or as non-supportive? Axial coding will then be conducted by which open codes are reviewed and synthesized. As a final step, selective coding will be conducted in order to articulate theories derived from the data and establish the inter-relations between the constructs that were found. By definition, Consensual Qualitative Research is conducted independently by two raters, who then meet to discuss their coding until a consensus is reached. A third independent rater will be available to address disagreements. All proposals are then edited by an editor.