Figure. Participant recruitment, randomization, and follow-up schedule.
Janda M, Neale RE, Youl P, Whiteman DC, Gordon L, Baade PD. Impact of a Video-Based Intervention to Improve the Prevalence of Skin Self-examination in Men 50 Years or OlderThe Randomized Skin Awareness Trial. Arch Dermatol. 2011;147(7):799-806. doi:10.1001/archdermatol.2011.48
Author Affiliations: School of Public Health, Institute of Health and Biomedical Innovation, Queensland University of Technology (Dr Janda), Cancer and Population Studies, Queensland Institute of Medical Research (Drs Neale, Whiteman, and Gordon), and Viertel Centre for Research in Cancer Control, Cancer Council Queensland (Ms Youl and Dr Baade), Brisbane, Australia.
Objective To improve the proportion of men 50 years or older who conduct a thorough whole-body skin self-examination (SSE), because these men are at increased risk of being diagnosed as having thick melanomas and dying of melanoma.
Design Randomized trial of a video-based intervention to increase whole-body SSE compared with brochures-only control.
Setting Men from an Australian population register.
Participants The study included 930 men 50 years or older.
Intervention The men were randomly assigned to receive either written materials only (control) or written materials along with a DVD or video and 2 postcard reminders (intervention). Telephone interviews were conducted at baseline and at 7 and 13 months.
Main Outcome Measure The proportion of men who conducted a whole-body SSE with the help of 2 mirrors and/or a person to assist with difficult-to-see areas.
Results Compared with baseline (10%), by 13 months both groups had increased their whole-body SSE behaviors similarly (36% intervention and 31% control) (P = .85). At 7 months, the intervention group was significantly more likely to examine at least 1 part of their back (P < .001) and to examine a greater number of body sites (P < .001). At 7 months, the proportion of men conducting any SSE increased by 28% in the intervention group compared with 13% in the control group (P < .001), but at 13 months, the prevalence of any self-examination (83% vs 80%) was again similar.
Conclusion While men 50 years or older are responsive to appropriately targeted intervention materials to increase their SSE behaviors, the addition of a video or DVD to written materials had only a transient advantage for optimal SSE practices.
Trial Registration anzctr.org.au Identifier: 12608000384358
In the United States, the incidence of melanoma in men 50 years or older has been steadily increasing over the last 2 decades, to 90 per 100 000 population in 2007, while remaining relatively stable in younger age groups.1 In Queensland, Australia, which has the highest rate of melanoma in the world,2 men 50 years or older have a 2-fold higher risk of melanoma compared with women of the same age (incidence rates, 209 and 112 per 100 000, respectively).2 Older men are also much more likely to be diagnosed as having thick melanomas and to die of melanoma.3 Death rates in the United States continue to increase among men diagnosed as having melanoma, while they are decreasing among women.4
There is now increasing evidence that a visual inspection of the skin by a physician (clinical skin examination [CSE])5 or a lay person (skin self-examination [SSE])6- 8 may result in reduced thickness of melanoma at diagnosis and thus improve survival in persons with melanoma. For example, in a Queensland case-control study, persons who received a whole-body CSE in the 3 years before their diagnosis had a 14% lower risk of being diagnosed as having a thick melanoma than those who did not, resulting in an estimated 26% fewer melanoma deaths in screened cases within 5 years in Queensland.5 A case-control study by Berwick et al6 showed that people who had ever performed SSE had a 63% lower risk of being diagnosed as having advanced disease. Recently Pollitt et al8 showed that the thoroughness of the SSE was the best predictor of reduction of melanoma thickness. While the United States Preventive Medicine task force cites insufficient evidence to recommend SSE for the general population,9 the American Cancer Society10 recommends monthly SSE, and the Cancer Council Australia recommends SSE every 3 months.11
Because melanoma frequently occurs on the back in men, it is important to check areas of the body that are difficult to see.12 The American Cancer Society recommends thorough SSE of all body areas, including the back, the back of the legs, and the scalp.10 However, the prevalence of such behavior is low in the general population. In a population-based telephone survey in Queensland, Australia,13 26% of participants self-examined the skin of their whole body at least once in the past year, but only 20% of men 50 years or older did so. A number of previous small trials successfully increased SSE in at-risk populations.14- 18 The large, population-based Check It Out trial defined thorough SSE as looking at all of 7 body areas at least once within the past 2 months.19 Using a video-based intervention plus a face-to-face counseling session and a telephone call by a health educator, the prevalence of thorough SSE increased from 18% to 55% over a 12-month period in the intervention group.20 However, while the Check It Out trial recruited persons attending physicians' practices, who would presumably already be health conscious, it did not specifically focus on men 50 years or older (mean [SD] age, 53.0 [14.8] years; 42% male), and it included resource-intensive intervention components.19 Given the success of the Check It Out trial, the importance of SSE in the absence of a routine population-based clinical screening program, and the increased risk of melanoma in men 50 years or older, we designed a randomized controlled trial to determine whether a video-based intervention without face-to-face contact could increase SSE behaviors specifically in men 50 years or older selected from the general population.
Our primary aim was to examine the impact of a video-delivered intervention with 2 mailed reminder postcards compared with a written-materials-only control group on the prevalence of whole-body SSE (wbSSE), intermediate SSE, and any SSE in men 50 years or older. The primary hypothesis of this trial was that the prevalence of wbSSE in the video intervention group would increase by at least 10% more than in the control group. A secondary hypothesis was that the intervention would be more effective in a subgroup of men with no previous treatment for skin lesions as well as in those who did not have a CSE in the past.
Trial methods and baseline participant characteristics have previously been reported in detail.21 Ethical clearance was provided by the Human Research Ethics Committee of the Queensland University of Technology, Brisbane, Australia, and the trial was registered with the Australian New Zealand Clinical Trials Registry.
In total, 5000 potential participants (men 50 years or older) were randomly selected from the Australian electoral roll (enrolling to vote is compulsory in Australia). Names and addresses were manually cross-referenced with the telephone directory, and those whose telephone number could not be found were excluded, resulting in a final list of 2899 potential participants. These men were mailed a study package, including a letter of invitation and a colored brochure featuring a well-known sports and TV personality endorsing the project. We sent 1 postal reminder and made up to 2 follow-up telephone calls to nonrespondents. Men were considered ineligible if they had a disconnected telephone line, were too ill, could not speak English, or had previously had a melanoma. The overall consent rate was 37% (969 of 2610 eligible); however, 39 men withdrew before the study began, leaving a final sample of 930 men (929 with baseline data) who were randomized to the control or the intervention group (Figure). Randomization was performed by computer-generated random number list, stratified by men's region of residence (metropolitan or other).
All participants received a package of study materials via mail approximately 2 weeks after their enrollment in the study. Those in the intervention group received a 12-minute DVD,21 featuring the above-mentioned sport and TV personality as narrator (for a detailed description of the DVD components, see reference 21). The DVD was developed after extensive pilot work, as described in our previous publication, and was guided by the extended Health Belief Model.22 The intervention group was also sent a body chart diagram (for men to record any suspicious skin lesions they found, to facilitate self-monitoring, and/or to aid recall when visiting a physician) that was suitable to be mounted on a refrigerator and a page of instructions on how to conduct a wbSSE, developed by the researchers (M.J., R.E.N., P.Y., and P.D.B.), titled The Ten-Step Guide to Checking Your Own Skin (available from the authors on request). Participants also received a standard color brochure issued by the Cancer Council Queensland titled Take the Time to Spot the Difference,23 showing benign and malignant lesions and identifying the features of each. Finally, 2 and 4 weeks after the initial mail-out, we sent the men a postcard reminding them to watch the DVD and to examine their own skin. Men in the control group received the researchers' guide and the standard color brochure but no video, body chart, or reminder postcards.
Computer-assisted telephone interviews were conducted by a professional survey company at baseline (before randomization) and at 7 and 13 months after study materials were sent to participants. At baseline, information was collected about the participants' hair, eye, and skin color before tanning; the degree of burning if their skin was exposed unprotected to strong summer sun for 30 minutes; their tanning response after being exposed to the sun for a number of days; the degree of freckling and moliness; whether they had made an attempt to get a tan in the past 12 months; whether they had been sunburned over the past 12 months; how frequently they used various measures of sun protection (rarely to always); whether they had ever had a skin cancer, mole, or other spot(s) removed or treated in the past; and whether they were currently concerned about a spot or mole. The participants were also asked whether they had undergone a skin examination by a physician in the past 12 months. (For a detailed description of all baseline data collected, see reference 21.) At each of the 3 time points, we collected information about the main SSE outcome measures.
Past SSE was assessed with the following 2 questions, which had been used previously24: “In the past 6 months, have you or someone who is not a physician, such as your spouse or partner, deliberately checked any part of your skin for early signs of skin cancer? ” “How often in the past 6 months have you or someone who is not a physician, such as your spouse or partner, checked your skin? ” The completeness of SSE was measured by asking the men to nominate which of 13 specific body areas they had examined during their last SSE (visible sites on the front of the body from the waist up and the front of the legs, as well as the back, buttocks, back of the legs, and back of the neck and scalp) and by asking whether they had used a mirror or another person to help them examine difficult-to-see areas.
We computed 3 measures of SSE, increasing in completeness: any SSE, intermediate SSE, and wbSSE. All men who indicated that they had deliberately checked at least 1 part of their skin in the past 6 months were considered to have performed any SSE. If the SSE included at least 1 part of their back during their last SSE (upper or lower back area, back of neck/scalp, or back of legs), they were also categorized as having performed intermediate SSE. Finally, the men were also categorized as having performed a wbSSE if they indicated that during their last SSE they had checked each of the 13 body areas mentioned and had either had another person help them check their skin or had used a full-length mirror and another mirror to check their skin. Other measures included the frequency of performing any SSE, the number of body areas checked during the last SSE, and, for those in the intervention group, the number of times they watched the intervention DVD.
Sample size calculations were based on our previous work indicating an approximately 20% baseline prevalence of wbSSE in men 50 years or older.13 Assuming a 10% attrition of participants, using a .05 (2-sided) significance level, a sample size of 500 men in each of the intervention and the control groups was calculated to provide 91% power to detect an increase of approximately 10% in the prevalence of wbSSE at 13 months. Power for subgroup analyses for equal subgroups (250 in each group) using the above assumptions was estimated at 80% to detect an increase of 12% in reported SSE at 13 months.
All statistical analysis was carried out according to the intention-to-treat principles, as specified in the trial protocol, using SPSS for Windows version 188.8.131.52 We used descriptive statistics to summarize the intervention and the control group participants' baseline demographic characteristics and SSE outcome measures at baseline and at 7 and 13 months. Loss to follow-up was low (10%).
We used binary logistic regression to assess the effect of the intervention on the main outcome measures at each time point. Generalized estimating equations were used to assess the effect over time. Each group's baseline SSE prevalence was used as the referent group for comparison with prevalence at subsequent time points. Given the very balanced distribution of the groups at baseline, there was no adjustment for potential confounders. To test our secondary hypotheses, we stratified the analyses by whether or not men reported having had a CSE in the past 12 months and whether they had ever had a skin spot or mole treated (as measured at baseline).
The demographic characteristics of the participants were similar between the intervention and the control groups, as previously reported21 and as shown in Table 1. Overall, to estimate men's health-related behaviors, 716 men (77%) had previously undergone a prostate-specific antigen test, and 366 (39%) had undergone a fecal occult blood test (Table 1). In total, 93% and 90% of the intervention group and 94% and 89% of the control group completed their follow-up telephone interviews at 7 and 13 months, respectively (Figure). Men who did not complete the 7-month interview (n = 60) and the 13-month interview (n = 99) were similar in terms of age, income, education, skin type, and past SSE compared with those who did. They were significantly less likely to have private health insurance than men who were surveyed at both time points (30% vs 47%; P = .02).
At baseline, a similar proportion of the intervention and the control groups had performed a wbSSE (12% vs 10%; P = .53), an intermediate SSE (47% vs 47%; P = .23), and any SSE (52% vs 56%; P = .19) (Table 2). Also, 23% of both groups had performed any SSE 3 times or more during the last 6 months.
The prevalence of wbSSE increased from baseline to 7 months in both the intervention (16% increase) and the control (13% increase) groups and by 7 months was similar between the groups (28% vs 23%; P = .11). By 7 months, a greater proportion of the intervention group than the control group had performed intermediate SSE (69% vs 57%, P < .001) and any SSE (80% vs 70%; P < .001). Similarly, significantly more intervention group participants reported performing 3 or more SSEs in the past 6 months (52% vs 43%; P = .009).
By 13 months, the prevalence of wbSSE had increased by a further 8% in both groups, resulting in 36% of the intervention group and 31% of the control group reporting performing a wbSSE in the past 6 months (P = .08). The intervention and the control groups reached similar levels of intermediate SSE (71% vs 68%; P = .32) and any SSE (83% vs 80%; P = .25). The prevalence of performing 3 or more SSEs during the past 6 months was also similar between groups (58% vs 55%; P = .32). After both follow-up time points in the generalized estimating equations models were taken into account, the group-by-time interaction was not significant for wbSSE (P = .85) but was significant for odds of intermediate SSE (P < .001) and any SSE (P < .001).
Compared with the number of body areas checked by the control group at baseline, the odds of examining a greater number of areas were higher for both groups at 7 months (intervention group: odds ratio [OR], 1.82; 95% confidence interval [CI], 1.32-2.31; control group: OR, 1.24; 95% CI, 1.01-1.53) and at 13 months (intervention group: OR, 2.57; 95% CI, 2.01-3.29; control group: OR, 2.19; 95% CI, 1.79-2.68) (P = .003).
Stratified prespecified analysis to test the secondary hypotheses showed that the intervention had a significantly larger effect in those men who had not undergone skin excision in the past (P = .03 for the 13-month comparison). However, SSE behavior by 13 months was similar in intervention and control group men who did not report a skin examination by a physician in the 12 months before baseline (P = .55) (Table 2).
Within the intervention group, we observed a significant dose-response effect of the DVD. Men who indicated that they had watched the DVD more than once in the past 6 months were significantly more likely to perform wbSSE at 7 and 13 months (P < .001) (Table 3).
At 7 months, a larger proportion of participants in the control group (78%) used the 10-step guide to SSE compared with the intervention group (71%; P = .03). At 13 months, a larger proportion of the control group (63%) compared with the intervention group (61%; P = .04) stated that they had used the color brochure. There was no other significant difference in the use of the written materials between the intervention and the control groups (Table 4).
Among all participants, the prevalence of wbSSE increased by approximately 20%, and any SSE increased by approximately 30% from baseline to 13 months; however, this increase was not notably different between the intervention and the control groups. The increased prevalence of SSE occurred sooner in the intervention group than in the control group. The video-delivered intervention was more effective than the control among men who had not previously had a skin lesion treated but not among men who had not had a CSE in the 12 months before the trial. Given that both intervention materials had a large effect, our results have important public health significance. It has been reported that conducting any SSE and increasing the thoroughness of SSE will increase the likelihood of presenting to a physician with a melanoma less than 1 mm thick,8 and reduction in thickness has the potential to improve survival after a diagnosis of melanoma among men 50 years or older.26
The difference between the intervention and the control groups in our study was not as large as that observed in the previous Check It Out trial, which included male and female participants of younger age on average than in our study.19 One reason for this difference could be that men assigned to the control group in our study received written information about skin cancer and SSE, while control group participants in the Check It Out trial received materials related to diet. Our control group appeared to use the written materials somewhat more than the intervention group. Older participants may prefer written information, which allows them to process information at their own pace. The additional multimedia information provided on the DVD may have been somewhat less successful in men 50 years or older than in the younger participants in the Check It Out trial. It has been highlighted that multimedia information may not enhance learning under all circumstances and may be a distraction for some learners.27
The written information may also have triggered men to discuss SSE with their physician. We have previously shown that a physician recommendation to perform SSE is one of the most important predictors of SSE.13,21 Written materials alone have been used in a number of previous studies involving older adults to improve health behaviors, with good success.28 However, the increase in SSE in the control group could also simply be attibutable to the participants' expectation of being asked about the behavior at a subsequent time or to participation in the study (Hawthorne effect).
We observed a strong dose-response effect, showing that watching the DVD more often was associated with higher odds of performing a wbSSE. The number of men who watched the DVD was greater in the first half of the intervention period than in the second half, and a subsequent reminder to watch the DVD again at 7 months may have boosted the intervention success and should be considered for later trials. Further analysis will explore in more detail the characteristics of those men who chose to watch the DVD more than once, as well as mediators of intervention effects in general.
As has been noted by Pollitt et al,8 different studies vary widely in the way that they define SSE as well as in the methods that they use to assess the completeness of SSE. In this study, we computed 3 different measures of SSE with increasing degree of completeness. Our definition of wbSSE meant that participants were required to be assisted by another person or to use 2 mirrors to view the back of the body. The video intervention provided a step-by-step demonstration of wbSSE, and men who received the DVD were more likely to check their back at the 7-month time point. Both groups received a written step-by step guide to wbSSE. At the end of the study, the control and the intervention groups were 3.7 and 4.3 times more likely (respectively) to have performed a wbSSE using a person to help or 2 mirrors to aid in checking the back compared with their respective baseline. While the proportion of participants who performed a wbSSE at the end of the study (36% of intervention group, 31% of control group) was lower than the 55% reported in the Check It Out trial, the Check It Out trial participants were not required to use these aids to be assigned a thorough wbSSE completion score.19 Pollitt and colleagues ’8 definition, however, required the participants to have used pictures of skin cancers to aid the wbSSE.
Video-delivered interventions have many obvious advantages, such as being appropriate to those with low literacy, being able to transmit culturally targeted clues, and being available for use at a time and location of the watcher's choice. While more intensive interventions or face-to-face interventions may be even more successful in increasing the prevalence of wbSSE, they may be too resource intensive to be disseminated population-wide. In contrast, distribution of our intervention seems feasible, and online distribution methods may further increase accessibility.29
The limitations of this study include reliance on self-report of SSE, which, when combined with data collection by telephone interviews, may have introduced social desirability bias. The nature of SSE behavior makes it impractical to measure by other methods such as direct observation. Compared with 2007 census data on Queensland men 55 years or older, a larger proportion of our participants were employed full-time (40% vs 26%), born in Australia (77% vs 65%), and had completed a university or college degree (23% vs 16%), indicating that we enrolled a somewhat more educated and affluent group of men. Participants in the skin awareness trial may have also been more health conscious than men from the general population. For example, a larger proportion of our participants reported that they had undergone either a fecal occult blood test or a prostate-specific antigen test (39% and 77%, respectively) compared with men in a previous cross-sectional telephone survey (16% and 52%, respectively).30 However, our data were obtained 5 years later than those reported by Carri ère et al,30 and the prevalence of a prostate-specific antigen test and a fecal occult blood test may have truly increased among the older population in Queensland. Because of the nature of telephone assessment, participants were excluded if they had no publicly listed telephone number, which may have limited the generalizability of our results. Given the increase in SSE behavior in the control group, subsequent trials may consider a no-intervention control group to allow a greater difference to be observed. Last, but not least, is the possibility that any additional benefit of a video-based intervention over written materials on wbSSE is smaller than we had predicted and would require an even larger trial to measure the true effect.
In summary, while the trial failed to reach a 10% difference in wbSSE between the intervention and the control groups at 13 months, the results are promising for the impact of interventions on improved SSE behavior among men 50 years or older. Significant increases in reported SSE prevalence were observed in both the control and the intervention groups. The addition of the DVD to the intervention materials demonstrated some advantages, especially at the 7-month time point, indicating quicker uptake, inclusion of crucial areas of the back in SSE, and greater increase in frequency of SSE compared with the brochure-only group. However, the advantages of inclusion of a DVD were not as pronounced as we would have expected based on previous studies and thus may be attributed to the more intensive face-to-face counseling also provided in those previous trials. Given reduction in mortality that can potentially be achieved by earlier detection of melanomas, providing support to men 50 years or older through appropriately designed written or written plus DVD intervention materials should be considered for population-wide dissemination, especially in high melanoma environments such as Queensland.
Correspondence: Monika Janda, PhD, School of Public Health, Institute of Health and Biomedical Innovation, Queensland University of Technology, Victoria Park Road, Kelvin Grove, Brisbane, Australia 4059 (firstname.lastname@example.org).
Accepted for Publication: December 22, 2010.
Published Online: March 21, 2011. doi:10.1001/archdermatol.2011.48
Author Contributions: Dr Janda had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Janda, Neale, Youl, Whiteman, Gordon, and Baade. Acquisition of data: Janda. Analysis and interpretation of data: Janda, Neale, Youl, Whiteman, and Baade. Drafting of the manuscript: Janda. Critical revision of the manuscript for important intellectual content: Janda, Neale, Youl, Whiteman, Gordon, and Baade. Statistical analysis: Baade. Obtained funding: Janda, Neale, Youl, and Baade. Administrative, technical, or material support: Janda and Youl. Study supervision: Janda.
Financial Disclosure: None reported.
Funding/Support: This study was funded by National Health and Medical Research Council (NHMRC) project grant 497200 and supported in part through funding provided to Dr Janda (NHMRC Career Development Award CDA 553034), Dr Neale (NHMRC Career Development Award CDA 552404), and Dr Whiteman (Australian Research Council Future Fellowship grant FT0990987).
Role of the Sponsors: The sponsors had no role in the design and conduct of the study; in the collection, analysis, and interpretation of data; or in the preparation, review, or approval of the manuscript.
Additional Contributions: Josephine Auster and Linda Finch assisted with project management. Lee Jones (BStats) assisted with data analysis. Joanne Aitken contributed to the design of the study and provided critical review of earlier versions of the manuscript. Ian Healy, Mark Smithers, Robert Jackson, and members of Melanoma Patients Australia generously gave their time and energy during the development of the intervention materials for this trial.