What motivates men age ≥ 50 years to participate in a screening program for melanoma?

Abstract

BACKGROUND.

The screening behavior and screening outcomes of men age ≥50 years was investigated within a randomized controlled trial of a community-based intervention of screening for melanoma, consisting of a community education program, an education program for medical practitioners, and the provision of dedicated skin-screening clinics.

METHODS.

Data from cross-sectional telephone surveys before (559 completed interviews), at the end (591 completed interview), and at 2 years after the intervention (445 completed interviews) were analyzed. In addition, the authors analyzed data from skin-screening clinics within the intervention program (3355 men age ≥50 years participated).

RESULTS.

During the intervention period men age ≥50 years increased both their screening behavior and intention to screen. Those men age ≥50 years who reported a past history of removal of a mole as well as other risk factors for skin cancer and positive attitudes toward screening were more likely to participate in skin screening across time. Men age ≥50 years accounted for 20.5% of all skin-screening clinic attendees, 31.3% of those referred for a suspicious lesion, 48.5% of melanomas, and 45% of all keratinocyte carcinomas diagnosed within the screening program, respectively.

CONCLUSIONS.

The intervention program successfully motivated men age ≥50 years to attend screening for skin cancer, resulting in the highest yield of skin cancer within this subgroup of the population. Messages addressing skin cancer risk factors and attitudes toward skin cancer and screening could be used to target a screening program for melanoma toward men age ≥50 years. Cancer 2006. © 2006 American Cancer Society.

Excluding keratinocyte carcinoma (KC; basal and squamous cell carcinoma), melanoma is the fourth most common cancer in Australia, with 8885 new cases diagnosed in Australia in 2001 (population of approximately 19,400,000).1 In the U.S., in which the incidence of melanoma is rising rapidly, 59,580 new cases of melanoma are expected for 2005 (population of approximately 294,000,000).2 Men accounted for 56% of all new melanoma cases, and for 63% of all melanoma-related deaths in both Australia and the U.S.1–3 The melanoma incidence continues to rise over-proportionally in men age >50 years.4

The thickness of the lesion at diagnosis is one of the most important prognostic indicators for survival from melanoma5, 6 and diagnosis and treatment of melanomas while they are still thin is likely to improve survival from this disease.7, 8 Screening for melanoma has the potential to improve early diagnosis. Although there is at present no conclusive evidence that screening for melanoma will reduce morbidity and mortality from melanoma, the U.S. Preventive Task Force (USPTF) describes screening as the most promising strategy, especially for older people.9 The American Cancer Society recommends a skin examination as a component of any routine cancer-related check-up.10 Targeted screening toward those age ≥50 years has been suggested as a possible way to increase its cost-effectiveness.11, 12 Previous studies found men age ≥50 years more commonly present with thick and nodular melanomas compared with women and younger people, and should be targeted by early detection programs.4, 13–18 Despite this, within general practice, excisions are more commonly performed on patients age <50 years compared with patients age ≥50 years.19 Older men are also less likely to self-present with a lesion of concern at open access community screening programs.20 However, despite representing only 25% of all screenees, men age >50 years contribute 44% of those patients with a confirmed melanoma within such clinics.21 A reduced ability to recognize a melanoma was reported for older compared with younger people, and older men frequently have lesions in difficult-to-see areas such as the scalp and the back, and therefore also may be limited in their ability to notice any new or changing lesions themselves.15

The present investigation presents an important part of the first phase of a randomized controlled trial of a community-based intervention of screening for melanoma.22, 23 Earlier, we reported the results from the baseline survey of this trial. People age ≥50 years were less likely to conduct a whole-body skin self-examination compared with younger people,24 but there was no difference noted between men and women or between different age groups with regard to self-reported prevalence of a whole-body skin examination by a general practitioner (clinical skin examination) within the past 12 months.25 Men were less likely to indicate an intention to conduct skin self-examination or to visit a physician for a skin examination within the next 12 months compared with women, suggesting that changing this behavior in men could prove challenging. Although having had a previous clinical examination by a physician was found to be most strongly related to future screening intention, several attitudinal factors (perceived susceptibility, giving skin checks a high priority) as well as a previous history of KC were also associated with intention to screen.26

After completion of the trial, we investigated changes in skin-screening behavior over time in intervention and control communities. Overall, within the intervention communities, the prevalence of whole-body skin examinations increased from 11% at baseline to a maximum of 34.8% 2 years into the trial, whereas screening rates among control communities remained stable. Uptake was highest among the population age ≥50 years.27

The current study investigated predictors of skin-screening participation of men age ≥50 years within the 3-year community-based screening program for melanoma. We also describe the clinical and histopathologic outcomes of screening examinations in men age ≥50 years who attended 1 of the dedicated skin-screening clinics provided during the intervention period.

MATERIALS AND METHODS

The design, intervention, and implementation of the intervention program has been described in detail elsewhere.22, 23 The aim of the trial was to determine the effectiveness of a community-based melanoma screening program in reducing melanoma mortality. The intervention consisted of 3 interrelated components: community education about early detection of skin cancer, education for local doctors in early detection of skin cancer, and dedicated skin-screening clinics. The objective was to increase to 60% the proportion of the population over 30 years within intervention communities who had at least 1 whole-body skin examination within the 3-year intervention period. During the first phase, 18 Queensland communities, each with an adult (age ≥30 years) population of ≥2000 were enrolled, for a total adult population of 63,035. Nine intervention communities were randomly allocated to receive a 3-year community-based melanoma screening program, with the remaining 9 control communities receiving standard practice only. The target population for screening was defined as those age ≥30 years, because mortality from melanoma is rare in patients younger than this age.1 For the purpose of this analysis, only data from the 9 intervention communities were used.

The respondents' self-reported skin-screening behavior, skin cancer risk factors, and attitudes toward skin cancer and skin screening were monitored by cross-sectional telephone surveys at 3 time-points: baseline (4 weeks before the intervention program during 1998), 36-month follow-up (at the end of the intervention period during 2001), and 5-year follow-up (2 years after the end of the intervention period during 2003).27

Telephone Surveys

Professional telephone interviewers used a Computer-Assisted Telephone Interview System (CATI) to reach random samples of community residents age ≥30 years selected from a commercially available directory of telephone numbers (equal numbers of men and women were ascertained through a quota system). The response rate for the 1998 survey was 66.9% (3110 completed interviews), 66.5% for the 2001 survey (5048 completed interviews), and 65.1% in 2003 (3514 completed interviews) (66.2% overall). Compared with the 1996 and 2001 Australian census, the sociodemographic characteristics of the survey respondents were similar with respect to sex, age, employment, and marital status in the respective populations. For the purpose of this investigation, only the data from men age ≥50 years were utilized.

Skin-Screening Clinics

During the intervention period, local doctors (in primary care practices) screened patients within their day-to-day practices and some also organized screening clinics within their private practices. In addition, centrally organized skin-screening clinics staffed by local doctors and additional doctors hired from outside the communities were held in intervention communities. Overall, within those dedicated clinics primary care physicians conducted 16,383 whole-body skin examinations. The aims, procedures, and outcomes of the skin-screening clinics have been described in detail elsewhere.27 Members of intervention communities were alerted to the clinics through advertisements and articles in local newspapers, street banners, flyers, and information brochures in physician practices and other places of interest, and personal letters of invitation signed by a sports celebrity for men and a media celebrity for women. The letters contained information regarding the time and place of the clinics and a toll-free telephone number for appointments. Care was taken to test the design of the letters directed at men for their suitability for this target group. The value of these letters for motivating men to attend screening with or without the addition of a glossy brochure was tested within a nested randomized trial at the beginning of the intervention period. There was no additional effect over and above the letter for the brochures and therefore the brochure was not used during subsequent invitations directed at men.28

Clinics were held in workplaces, community venues, and local hospitals and participants completed a questionnaire before their examination that included demographic characteristics, history of skin screening, and skin cancer risk factors. Participants signed consent forms for the whole-body examination and to allow access to relevant medical information resulting from the skin examination. Overall, 2302 clinic attendees (14.1%) were referred back to their own physician for investigation of 4129 suspicious lesions. The location and provisional diagnosis of the suspicious lesion discovered during the examination was noted by the skin clinic physician on the referral form, which was given to the patient. A copy was also retained by the research team. Of those referred, 1822 (79.2%) filled their referral after a maximum of 2 telephone and written reminders. For these patients, their physician recorded how the lesion was managed and the date of management. If a lesion was excised or biopsied, histopathology reports were obtained.

Measures

A detailed description of the telephone survey development and psychometric testing can be found elsewhere.25, 26 The following variables were measured during the telephone interviews and used for the present analysis: 1) sociodemographic factors (sex and age), 2) skin cancer risk factors (history of spot or mole removal, personal and family history of KC or melanoma), 3) attitudes and intentions toward skin screening (intention to have a clinical skin examination within the next 12 months, intention to examine their own skin within the next 12 months, perceived susceptibility to develop skin cancer, concern about skin cancer, current concern about a specific spot or mole), and 4) past skin screening behavior (skin screening by whole-body skin examination by a physician during the past 12 months).

Data Analyses

Changes in reported screening behavior, attitudes, and intentions toward skin cancer screening based on the telephone surveys were assessed using logistic regression models. Separate models were used for each outcome. A 3-level categoric variable representing the survey period was the dependent variable, and baseline prevalence was taken as the reference point. The change in attitudes and behaviors over time was expressed as adjusted odds ratios (OR) and 95% confidence intervals (95% CIs). The significance of the overall change was assessed by the Wald chi-square statistic from the logistic model. To adjust for the cluster design of the study (with communities being the cluster unit), the logistic regression models were performed using the SUDAAN statistical package (Research Triangle Institute, Cary, NC),29 which allows for the increase in variation associated with this type of study design.

We also fitted multivariate random effect models (using the glimmix macro in SAS software [SAS Institute Inc., Cary, NC]) to assess the influence of physician recommendations or instructions to self-examine the skin, skin cancer risk factors, and attitudes regarding skin-screening participation throughout the trial period. Separate models were used for each factor and included 2-way interaction terms between time and the specific factor.

Ethics

The Behavioural and Social Sciences Ethical Review Committee of the University of Queensland approved this study.

RESULTS

Telephone Survey Results

Within intervention communities, men age ≥50 years were nearly 4 times more likely to report a clinical whole-body skin examination within the past 12 months at the end of the intervention (32.8%) compared with baseline (10.6%), and still more than twice as likely 2 years after completion of the intervention (24.8%). Men age >50 years were twice as likely to report conducting a whole-body skin self-examination within the past 12 months at the end of the intervention (27.5%) and at 2 years of follow-up (28.0%) compared with baseline (15.8%). In addition, men age ≥50 years were significantly more likely to report an intention to conduct skin self-examination at the end of the intervention (85.0%) and at the 2-year follow-up (81.2%) compared with baseline (60.8%) (Table 1).

Table 1. Changes in Skin Screening Behavior, Attitudes, and Intentions from Baseline to the End of the Intervention and 2 Years after the Intervention among Men Age ≥50 Years (Intervention Communities Only)

	Odds ratio (95% CI)*			Wald F; P
	Baseline survey† (n = 559)	Survey end of intervention† (n = 591)	Survey 2-year follow-up† (n = 445)
95% CI indicates 95% confidence interval; KC, keratinocyte carcinoma (basal and squamous cell carcinoma). * The odds ratio (95% confidence intervals) was calculated by logistic regression models using SUDAAN adjusted for cluster sampling using the baseline survey as the reference category. † Cross-sectional survey at each time point.
Had a clinical whole-body skin examination in the past 12 months	1.00	4.12 (2.62–6.47)	2.78 (1.73–4.47)	19.6; <.001
Has performed whole-body skin self-examination within the past 12 months	1.00	2.01 (1.29–3.14)	2.06 (1.31–3.26)	5.30; .005
Intend to have a clinical skin examination within the next 12 months	1.00	1.13 (0.74–1.72)	1.03 (0.67–1.58)	0.33; .72
Intend to check my own skin in the next 12 months	1.00	2.14 (1.30–3.53)	1.62 (0.98–2.69)	4.74; <.001
Doctor suggested to perform skin self-examination within the past 12 months	1.00	0.95 (0.52–1.72)	1.04 (0.57–1.91)	0.17; .85
Doctor taught how to perform skin self-examination within the past 12 months	1.00	1.74 (0.98–3.08)	1.54 (0.86–2.78)	1.81; .16
Very likely to develop skin cancer	1.00	0.60 (0.39–0.92)	0.67 (0.43–1.04)	2.69; .07
Current concern about a spot or mole	1.00	0.98 (0.59–1.63)	0.70 (0.41–1.20)	2.10; .12
Very concerned about skin cancer	1.00	0.78 (0.50–1.23)	0.73 (0.45–1.16)	0.89; .41
Regular skin checks are a priority	1.00	0.84 (0.53–1.33)	0.74 (0.46–1.20)	0.89; .41
Confident that I could find a suspicious spot or mole	1.00	0.63 (0.38–1.06)	0.43 (0.25–0.73)	6.08; .002
I would contact the doctor immediately if I found something suspicious	1.00	1.43 (0.86–2.39)	1.42 (0.83–2.41)	1.00; .37
Confident that doctor can diagnose skin cancer	1.00	0.96 (0.55–1.66)	1.01 (0.57–1.77)	0.04; .95
Ever had a spot or mole removed	1.00	0.94 (0.60–1.46)	0.89 (0.56–1.40)	0.16; .85
History of melanoma	1.00	1.07 (0.54–2.12)	1.46 (0.72–2.93)	0.99; .37
History of KC	1.00	1.30 (0.74–2.30)	1.49 (0.83–2.67)	1.00; .37

At the end of the intervention period (22.5%) and 2-year follow-up (24.2%), we found no increase in the proportion of men age ≥50 years who indicated that their physician suggested they conduct a skin self-examination compared with baseline (23.4%). Somewhat more men indicated at later timepoints (15.7% and 14.2%, respectively, at end of the intervention and at 2 years of follow-up) that they received instructions from their physician regarding how to conduct a skin self-examination compared with baseline (9.6%); however, this did not reach statistical significance as indicated by the 95% CIs (Table 1). Similarly, the proportion of men who perceived themselves to be at high risk of skin cancer reported a current concern about a spot or mole, were very concerned about skin cancer, or expressed the view that skin checks are a priority remained stable throughout the intervention period. Men age ≥50 years reported a significant reduction in confidence that they could find a suspicious spot or mole at the end of the intervention (79.4%) and at 2 years of follow-up (72.5%) compared with baseline (85.9%). There was no significant change noted in the participants' attitude toward contacting a physician if they detected something suspicious on their skin, or confidence in their physician's ability to diagnose skin cancer throughout the observation period. There was a slight decrease in the number of men age >50 years reporting that they ever had a spot or mole removed in the past (67.4% at baseline; 65.9% at the end of the intervention; and 64.7% at 2 years of follow-up), or reporting a history of melanoma or KC compared with baseline (Table 1).

The random effects models to explain the influence of skin cancer risk factors and attitudinal factors concerning participation in skin screening across all 3 timepoints revealed a significant interaction effect for the models “time” × “doctor taught how to perform a skin self-examination” (F(2;1394) = 6.52; P ≤ 0.001), “time” × “very likely to develop skin cancer” (F(2;31565) = 9.77; P ≤ 0.001), and “time” × “confident that I could find a suspicious spot” (F(2;1565) = 3.13; P = 0.04) (Table 2). Generally, the immediate effect of the intervention was to reduce the effect of these variables on the outcome measure. For example, at baseline nearly twice as many men who thought they were very likely to develop skin cancer reported having recently received a whole-body skin examination (14%) compared with those who thought they were unlikely (8%). After the intervention, the difference in reported whole-body skin examinations was reduced (from 34% to 32%). This reduction in these subgroup differences did not persist long term (i.e., 2 years after the intervention), particularly for “doctor-taught skin examination” and a perception of increased personal risk of skin cancer. All other interactions were not significant, indicating that men reporting these skin cancer risk factors or positive attitudes toward skin screening were consistently more likely to also report a whole-body skin examination within the past 12 months across all 3 timepoints (Table 2).

Table 2. Influence of Attitudes, Skin Cancer Risk Factors and Doctors Recommendation of Skin Self-Examination on Whole-Body Skin Examinations

		% (n) Reporting a whole-body skin examination within the past 12 months			F-value (interaction); P
		Baseline survey* (n = 559)	Survey end of intervention* (n = 591)	Survey 2-year follow-up* (n = 445)
Y indicates yes; N, no; KC, keratinocyte carcinoma (basal and squamous cell carcinoma). * Cross-sectional surveys at each time point.
Doctor suggested to perform skin self-examination within the past 12 months	Y	31.2 (36)	58.0 (68)	53.4 (49)	1.97, .14
	N	6.3 (39)	23.7 (96)	13.7 (41)
Doctor taught how to perform skin self-examination within the past 12 months	Y	66.1 (24)	66.3 (51)	60.9 (35)	6.52, .001
	N	5.19 (51)	24.8 (113)	16.6 (56)
Very likely to develop skin cancer	Y	14.1 (33)	34.1 (61)	41.7 (61)	9.77, <.001
	N	8.1 (44)	32.0 (135)	16.7 (48)
Current concern about a spot or mole	Y	13.4 (20)	28.3 (35)	31.2 (22)	2.07, .13
	N	9.9 (57)	33.7 (161)	23.6 (87)
Very concerned about skin cancer	Y	15.7 (38)	28.3 (35)	31.2 (22)	1.98, .14
	N	9.9 (57)	33.7 (161)	23.6 (87)
Regular skin checks are a priority	Y	14.9 (68)	40.2 (136)	31.2 (73)	0.89, .41
	N	3.9 (9)	22.7 (59)	17.4 (36)
Confident that I could find a suspicious spot or mole	Y	9.8 (65)	31.3 (147)	26.6 (84)	3.13, .04
	N	15.6 (12)	37.7 (49)	19.9 (25)
I would contact the doctor immediately if I found something suspicious	Y	13.3 (74)	35.9 (180)	26.8 (84)	1.58, .21
	N	1.11 (3)	16.1 (16)	14.7 (11)
Confident that doctor can diagnose skin cancer	Y	11.5 (70)	33.5 (165)	25.6 (95)	0.53, .59
	N	6.2 (7)	28.4 (31)	20.5 (14)
Ever had a spot or mole removed	Y	10.5 (61)	37.7 (145)	28.5 (86)	0.57, .55
	N	10.9 (16)	24.2 (51)	17.9 (23)
History of melanoma	Y	20.3 (9)	48.6 (20)	41.2 (16)	0.31, .73
	N	10.0 (68)	31.5 (176)	23.2 (93)
History of KC	Y	19.8 (12)	44.3 (49)	39.5 (37)	0.57, .57
	N	7.6 (33)	30.1 (147)	21.0 (72)

Skin-Screening Clinic Outcomes for Men Age ≥50 Years

Men age ≥50 years comprised 22% of the population within intervention communities, and 20.5% of all those who attended the skin-screening clinics provided as 1 component of the intervention program; however, this same group comprised 31.3% of those referred for a suspicious lesion. Of those men age ≥50 years referred (n = 720), 81.5% visited a physician for their referral. Of all 1343 lesions excised or biopsied, 31.3% were detected in men age ≥50 years and, of the lesions assumed to be melanoma, 32.7% were detected in men age ≥50 years. Of the histologically confirmed melanomas, 48.5% were found within this same group. Of all basal cell carcinomas and squamous cell carcinomas diagnosed within the screening program, 46.3% and 44.3%, respectively, were found on men age ≥50 years.

The overall yield of melanoma in the screening program was 33 of 16,383 (2.0 per 1000 screenings), with a yield of 16 of 3355 (4.8 per 1000 screenings) in men age ≥50 years compared with 17 of 13,028 among all other participants (1.3 per 1000 screenings). Of the 16 melanomas detected in men age ≥50 years, 50% were in situ and 50% were <1 mm thick. The most common location of melanomas in men age >50 years was on the back (69%). Of men age ≥50 years who were diagnosed with melanoma during the screening program, all reported a past history of having a spot or mole removed, and 11 men (68.8%) were currently concerned about a spot or mole.

DISCUSSION

The results of the current study indicate that within a community-based randomized screening program, men age ≥50 years can be motivated to participate in screening for melanoma and that with screening examinations, melanomas are more commonly detected in men age ≥50 years than other screening participants. Despite the fact that men indicated less intention to screen for melanoma compared with women at baseline,26 the community intervention program was successful in inducing behavior change in men age ≥50 years by increasing their rate of whole-body clinical skin examinations by 4-fold, and the rate of skin self-examinations by 2-fold. Also, their screening behavior during the 2 years after the intervention declined more slowly compared with screening participants overall.27

The community-based intervention program of melanoma screening employed within this randomized trial was not specifically targeted at men age ≥50 years. However, particular care was taken to ensure it was also suitable to a male audience.23 All materials were written at primary school reading levels, and earlier we reported that residents of all educational levels were equally likely to attend the screening clinics.30 Skin-screening clinics were advertised in local newspapers and all residents were sent letters of invitation that were thoroughly pilot tested.28 Screening services were located in workplaces, hospitals, and community centers to make them easily accessible for men. In addition, the educational package for medical practitioners included information concerning the disproportional risk of thick melanomas in men age ≥50 years. All these components may have contributed to the successful behavioral change observed among men age ≥50 years within the current study.

Across all time periods, those men who reported risk factors such as the removal of a spot or mole in the past or current concern about a spot or mole were consistently more likely to participate in screening compared with men age ≥50 years without such a history, and all men diagnosed with melanoma reported the removal of a spot or mole in the past. These results are similar to findings from the American Academy of Dermatology (AAD) National Skin Screening Program, resulting in the recommendation to focus screening toward men with a changing mole to further increase the yield of melanoma within the screened population.21 These results also support the notion that the natural increase in removal of benign lesions and KCs through a screening program for melanoma may positively influence screening participation during subsequent rounds by increasing awareness of the importance of skin checks.31 Conversely, future screening programs will need to target those men age ≥50 years who never had a spot or mole removed and who are unconcerned about a mole or spot to raise their attendance rates.

The association between confidence to self-detect a spot and screening behavior under the influence of the intervention program was found to be somewhat different from the other attitudinal factors. Men who reported less confidence in their own ability to self-detect a spot or mole were more likely to report a skin examination at baseline and at the end of the intervention program, but fell below men with such confidence after the intervention ceased. Overall, we observed that men's confidence to self-detect a spot or mole of concern decreased significantly over time. This suggests that addressing men's confidence and self-detection skills could be important within skin-screening promotion, possibly through increasing skin self-examination recommendations and instructions by general practitioners. Within the current screening program, we did not observe a significant increase in physicians recommending such skin self-examination behavior, leaving quite a large leeway for improvement. In an innovative approach, which may be transferable to a population setting, high-risk patients were provided with images of their body surface and instructions to systematically examine the skin of the whole body.32 This intervention increased the median frequency of self-examination from twice yearly to 6 times yearly. Greater than 50% of participants used the images provided to assist their self-examination,32 thereby potentially increasing their confidence that they could detect any notable change.

Men age ≥50 years contributed 30% of the referred patients, and also 30% of the patients who underwent a biopsy or excision for a suspicious lesion. In addition, nearly 50% of the melanoma cases were detected on men age ≥50 years. These figures are very similar to those reported by the AAD National Skin Screening program, in which dermatologists screened >600,000 individuals over the past 15 years.20 Within this program, men age ≥50 years comprised 25% of all screening participants, but 44% of all confirmed melanoma cases. The yield per 1000 screenings was higher among men age >50 years within the current screening program (4.8) compared with the AAD program (2.6), highlighting the importance of screening efforts for men age ≥50 years within Australia.

Within this community-screening program, men age ≥50 also had the highest yield of KCs and seborrhoeic keratoses. Although the detection of KCs is not the primary aim of the program, removal of a KC or even a benign lesion could increase the willingness of men age ≥50 years to undergo further screening for melanoma through changes in perceived susceptibility.33 The time spent with the medical practitioner while undergoing an excision or biopsy may also provide an opportunity for health education in a teachable moment.24

The main health promotion messages of the community-based intervention program encouraged attendance at a skin-screening clinic for a whole-body screen by a physician, but also encouraged participants to examine their own skin and present to a physician with any suspicious lesion. In a case-control study of melanoma in Connecticut, skin awareness was associated with a favorable prognosis in addition to other, well-known predictive factors of melanoma outcome such as Breslow thickness and mitotic index.34 It has been suggested that men should be encouraged to note any change in size of a lesion and that this should trigger action even in the absence of other symptoms.16 However, in the current study, similar to previous observations, the majority of melanomas in these older men were located on the back—an area that is difficult to see.15 These findings also point toward the limitations of part-body examinations and self-examinations if these are not conducted thoroughly, using a mirror or utilizing the help of a second person to locate lesions on difficult-to-see areas.8

This randomized trial employed a nested cross-sectional design to monitor uptake of screening. The advantage of using such design is that the results are not affected by cohort movements (in-migration and out-migration). However, this design does not allow statistical modeling of changes in behavior and intentions to the same extent as a nested cohort design with longitudinal measurement.35 Although the yield of melanoma cases detected within the screening program was higher than reported from open skin-screening days in the U.S. or elsewhere,36 the number of melanomas detected is still small, and therefore the results need to be interpreted with caution. The generalizability of these results may be limited, as the study communities were located in rural and regional areas of Queensland. Although the participants had similar sociodemographic characteristics compared with the Australian census, they may have differed in some characteristics from those not participating in the surveys.

To our knowledge, the current study is the first to report on the screening behavior and detection patterns among men age ≥50 years within a population-based screening program for melanoma. The results provide evidence that the skin-screening behavior of this population subgroup is amenable to change through a community-based intervention program. However, the ability to sustain high levels of screening activity will depend on the availability of services. Future population-based melanoma screening programs need to emphasize the importance of whole-body examinations both by patients themselves and their physicians. Our results also suggest that to sustain screening rates in men age ≥50 years an understanding of their susceptibility to melanoma and their physician's encouragement of early detection and screening behavior will be important.