“Watch your back” melanoma educational campaign.
Geller AC, Johnson TM, Miller DR, Brooks KR, Layton CJ, Swetter SM. Factors Associated With Physician Discovery of Early Melanoma in Middle-aged and Older Men. Arch Dermatol. 2009;145(4):409–414. doi:10.1001/archdermatol.2009.8
To determine factors associated with physician discovery of early melanoma in middle-aged and older men.
Three institutional melanoma clinics.
A total of 227 male participants (aged ≥40 years) with invasive melanoma who completed surveys within 3 months of diagnosis.
Main Outcome Measures
Factors associated with physician-detected thin melanoma.
Patients with physician-detected melanoma were older, 57% were 65 years or older compared with 34% for other-detected (odds ratio [OR], 2.57; 95% confidence interval [CI], 1.19-5.55) and 42% for patient-detected melanoma (P = .07). Physician-detected melanoma in the oldest patients (aged ≥65 years) had tumor thickness equal to that of self-detected melanoma or melanoma detected by other means in younger patients. Back lesions composed 46% of all physician-detected melanoma, 57% of those detected by other means, and 16% of self-detected lesions (physician- vs self-detected: OR, 4.25; 95% CI, 1.96-9.23). Ninety-two percent of all physician-detected back-of-the-body melanomas were smaller than 2 mm compared with 63% of self-detected lesions (P = .004) and 76% of lesions detected by other means (P = .07).
Skin screenings of at-risk middle-aged and older American men can be integrated into the routine physical examination, with particular emphasis on hard-to-see areas, such as the back of the body. “Watch your back” professional education campaigns should be promoted by skin cancer advocacy organizations.
Less than a third of melanomas are first discovered by a physician, but physicians detect melanoma at an earlier stage than the patient or significant other, such as a spouse or partner.1- 7 There is limited understanding of the factors associated with discovery of melanoma and the variables related to early detection, especially in the high-risk group of middle-aged and older men. Public education recommendations have encouraged patients to request physician skin examinations, and professional educational messages have encouraged physicians to perform full-body skin examinations.8,9 Our objective was to characterize patient factors associated with physician discovery of early melanoma in middle-aged and older men.
Institutional review board approval for patient accrual was obtained at Stanford University Medical Center (SUMC), Stanford, California; Veterans Affairs Palo Alto Health Care System (VAPAHCS), Palo Alto, California; and the University of Michigan (UM), Ann Arbor. Following informed consent, men aged 40 years or older with a diagnosis of invasive primary cutaneous melanoma were surveyed in the tertiary melanoma clinics of these institutions from September 1, 2004, through January 31, 2006, as previously described.10 Study sample, tumor characteristics, and staging criteria have also been previously described.10 Patients with thicker (>2-mm) and thinner (≤2-mm) primary tumors were compared in terms of the following 4 categories: (1) histopathologic characteristics and anatomic location of the primary melanoma with particular attention to the back and other back-of-the-body sites (posterior neck, shoulders, buttocks, arm, thigh, calf), (2) patient characteristics (age, patient report of atypical nevi, tendency to sunburn), (3) patient awareness and attitudes (including awareness of melanoma, interest in their health, reading and watching news stories about health, recognizing the importance of the medical examination for skin cancer, attention paid to skin cancer information, awareness, and confidence in early mole detection), and (4) medical access and skin cancer examination practices (physician and patient skin examination practices). For most of the questions, respondents were asked to refer to the time period 1 year prior to diagnosis.
We sought to determine associations of these factors with the discovery pattern, namely, who discovered the melanoma. The “discoverer” was either the physician, the patient (self-detected), or other (including the spouse, significant other, or partner; family member; or friend; hereinafter “other detected”). The spouse or significant other or partner (as opposed to another family member or friend) discovered all but 10 of the other-detected melanomas. Most of the results were analyzed as a 3-group sample (dermatologist or other health care provider, including nondermatologist physician vs other vs self). Differences were tested by χ2 analysis. A critical test value of P < .05 was used throughout the analysis.
Of the 227 melanoma diagnoses, 170 (75%) had a Breslow depth of 2 mm or less, and 57 (25%) were thicker (>2 mm). Melanoma was first discovered by the physician in 58 men, by the patient himself in 84 cases, and by others in 53 cases (Table 1). Among men who had at least 1 physician visit in the year before diagnosis, 48% performed skin self-examination, 41% received a full-body skin examination (including head, trunk, arms, legs, hands, and feet), and 26% asked a physician for a skin examination.
Among physician-detected melanomas, 69% had a thickness of 1 mm or less (T1), compared with 50% of those detected by others (P = .04) and 33% detected by the patient (P < .001) (Table 1). For physician-detected melanomas, 10% were later stage (American Joint Committee on Cancer stages IIA, IIB, IIC) compared with 28% detected by others (P = .02), and 35% of self-detected melanomas (P = .001). For melanomas first detected by a physician, 78% were superficial spreading or lentigo maligna melanoma subtypes, similar to histologic subtypes detected by others (76%) but in a significantly greater proportion than in self-detected melanomas (51%; P = .001), which had an increased proportion of the nodular subtype. Three percent of physician-detected melanomas were ulcerated, compared with 20% of other-detected lesions (P = .005) and 20% of self-detected lesions (P = .005).
Fifty-seven percent of men with physician-detected melanoma were 65 years or older compared with 34% of other-detected lesions (odds ratio [OR], 2.57; 95% confidence interval [CI], 1.19-5.55) and 42% of patient-detected lesions (P = .07). Compared with men younger than 65 years, older patients were more likely to have had at least 2 medical visits in the year prior to melanoma diagnosis (83% vs 66%) (P = .005). Among men of all ages, 81% with a physician-detected melanoma had made at least 2 physician visits in the past year, significantly more than men with a melanoma detected by others (60%; P = .02) (OR, 2.08; 95% CI, 1.19-6.60) but equal to men with self-detected melanoma (78%) (Table 1).
Back lesions composed 46% of all physician-detected melanoma, 57% of other-detected, and 16% of self-detected melanoma (physician vs self-detected: OR, 4.25; 95% CI, 1.96-9.23). Lesions on any part of the back of the body composed 63% of physician-detected melanoma, 75% of other-detected melanoma, and 42% of self-detected melanoma (physician vs self-detected: OR: 2.40; 95% CI, 1.20-4.79; other vs self-detected: OR, 4.09, 95% CI, 1.91-8.79). Conversely, melanoma found on the arm was twice as likely to be detected by the patient as by a physician (34% vs 14%; P = .008) or patient vs other (34% vs 16%; P = .02). There were no differences in anatomic location between dermatologist- and nondermatologist-detected melanoma.
Physicians detected significantly thinner back-of-the body lesions than did patients, and there was a trend toward thinner melanomas compared with those detected by others. Ninety-two percent of all physician-detected back-of-the-body melanomas were smaller than 2 mm compared with 63% of self-detected melanomas (P = .004) and 76% of other-detected lesions (P = .07). There was a trend toward earlier-detected back lesions by physicians, but this was not statistically significant (P = .08). Conversely, there was little difference in tumor thickness for front-of-the-body melanomas whether detected by physicians, patients, or others.
Overall, prior to diagnosis, few men requested educational materials regarding skin cancer detection from their physicians but significantly more with physician-detected melanoma did compared with those with self-detected melanoma (11% vs 3%, respectively; P = .045). Men whose melanoma was first detected by their physician showed a trend toward having greater interest in reading or watching news stories about health or believing that it was important to have a physician examine their skin (Table 2). Similar proportions of men with physician- or self-detected melanoma reported a physician talk about melanoma before diagnosis (59% and 56%, respectively), higher rates than for men with other-detected melanoma (39%; P = .03 for physician vs others). Requests for skin cancer examinations or use of pictures or photographs to assist in looking at moles did not differ by who discovered the melanoma (Table 3).
For the 59 melanomas discovered by a physician, there were no differences between dermatologists (n = 39) and nondermatologist physicians (n = 20) in tumor thickness, stage of diagnosis, histologic characteristics, or ulceration of the primary tumor. The few discovery pattern differences between dermatologists and nondermatologists were evident in patient recall of being instructed or given materials on how to look at the skin (dermatologists, 54%, vs nondermatologist, 5%; P = .002), physician talk about melanoma (77% vs 25%; P = .001), and provider performance of a full-body skin examination (56% vs 26%; P = .04). Among patients with physician-detected melanoma, more who saw dermatologists stated that it was important to have a physician examine their skin for cancer (dermatologist, 87%, vs nondermatologist, 60%; P = .02).
Men with physician-detected melanoma had a tendency to have more clinically atypical nevi (38% vs 23% for other-detected melanoma; P = .09, and 38% vs 16% for self-detected melanoma; P = .003) (OR, 3.23; 95% CI, 1.45-7.23).
Men with a physician-detected melanoma were no more likely to report having a full-body skin examination in the year before diagnosis compared with patient- or other-detected melanoma. There were no differences in careful skin self-examinations among patients with physician-detected (43%), self-detected (41%), or other-detected melanoma (52%) (physician vs self; P = .28).
Surveillance of the skin during routine medical care could save many lives otherwise lost to melanoma.11,12 Visual examination by a qualified observer takes minimal time and is regarded by many as reliable in diagnostic situations.11,12 Such examinations could enhance detection of melanomas on the back of the body, including the back and posterior legs, which cannot be viewed as easily by patients. Patients with melanoma typically have contacts with physicians in the year before diagnosis, suggesting the potential for many lesions to be diagnosed earlier. In one study13 of 216 melanoma cases, 87% of the patients had regular physicians, 63% had seen those physicians in the year prior to diagnosis, but only 20% reported physician skin examinations and 24% reported skin self-examinations prior to diagnosis.
Melanoma is more frequently discovered by the patient or a significant other, but physician discovery is more commonly associated with thinner lesions.2- 7 Our study in middle-aged and older men is consistent with other studies that show that physicians routinely detect thinner melanoma than patients.2- 7
The purpose of this study was to first understand what factors predispose physicians to detect melanoma and then to explore whether any of these factors are related to earlier diagnosis. When compared with self-detected and other-detected melanoma, we found that men with physician-detected melanoma were more likely to have back or back-of-body lesions, more commonly report a history of clinically atypical nevi, and were more likely to be older than 65 years. However, contrary to our hypothesis, men with physician-detected melanoma were no more likely to report receiving a full-body skin examination than men with self- or other-detected melanoma.
Melanoma can present anywhere on the body, but in men the back is the most common site, making up nearly a third of all cases.13 Likely because men also find it hard to see these areas, back or back-of-the-body melanomas are generally thicker than those in other sites.14 Therefore, reducing thick back lesions in men should be a paramount melanoma control strategy. Our results show that back lesions made up nearly half of all physician-detected lesions, and back-of-the-body lesions comprised 63% of physician-detected melanoma. Notably, 92% of physician-detected back-of-body lesions were thinner than 2 mm, compared with 63% of self-detected and 76% of other-detected lesions.
Multiple studies show that patients with atypical nevi present with thinner melanomas compared with patients without atypical nevi.3,15,16 In our study, men with self-reported clinically atypical nevi were more likely than men without atypical nevi to have a physician-detected thin melanoma. This important risk factor may prompt patients to request physician examinations and physicians to more carefully follow at-risk patients.
Older age is generally associated with thicker, more invasive melanoma; however, older patients are more likely to have multiple physician visits,17 thereby providing numerous opportunities for a skin examination. Older patients may rely on a physician evaluation of the skin because they have more problems with their eyesight or fewer partners to examine their skin. Our data demonstrate the importance of physician examination for early detection because men 65 years or older whose melanoma was detected by their physician generally had thinner tumors, similar to younger patients with a self-detected or other-detected melanoma.
Epidemiologic and ecologic studies further support the physician's role in early detection of melanoma. Schwartz et al3 examined the characteristics of 1515 male and female patients and also found that physician detection and the presence of 1 atypical nevus were associated with thinner melanoma. Studies of physician supply in Florida indicated that among male patients, an increasing supply of family physicians was associated with higher melanoma incidence and lower melanoma mortality.18 Likewise, increasing supplies of dermatologists were associated with lower overall melanoma mortality rates.
Study results should be interpreted in light of several limitations, which have been previously described.10 Subtle differences in how subjects recall events leading up to discovery could result in misclassification of the “discoverer” of melanoma. For example, a patient might examine his or her skin, identify a suspicious mole, and ask the physician to examine the mole further. If the physician makes the final diagnosis of the melanoma that arose from the mole prompted by a patient request, patients may report these events in different ways. The methods of data collection and the cross-sectional nature of the study also limit our ability to examine many complex and subtle interactions between physicians and patients, such as patient requests for skin examinations, physician response to requests, and the thoroughness and quality of the examination. Our finding that 41% of patients received a full-body physician skin examination prior to diagnosis might be underestimated because patients might be unaware that their skin is being examined during routine medical visits. However, the relatively high rates of physician examinations found herein are double those recently found by Rodriguez et al19 of nondiseased patients attending primary care and dermatology clinics, leaving open the possibility that people with melanoma are more likely to recall earlier examinations.
There are a number of factors that were not explored in this study that might also promote earlier detection by physicians. Compared with patients and others, physicians are more likely to be aware of risky moles, use improved lighting for doing a skilled examination, and may have better opportunities to examine intermittently exposed, fully exposed, and/or unclothed sites. Once a suspicious mole is found, physicians can more effectively motivate patients to seek immediate follow-up care and facilitate referrals for biopsy to avoid patient delay.
Our findings have implications for promoting early treatment of melanoma and emphasize the importance of physician discovery. Although most patients had regular physicians, only a minority reported physician skin examinations or skin self-examinations. Boosting physician and self-screening rates for middle-aged and older men requires a 2-fold public and professional educational strategy. With appropriate training, skin screenings can be brief and integrated into the routine physical examination, with particular emphasis on hard-to-see areas such as the back. “Watch your back” (Figure) professional education campaigns should be promoted by skin cancer advocacy organizations and should incorporate the importance of physician screening10 and the benefit of spouse or partner assistance for early detection of melanoma,20 particularly in the high-risk population of middle-aged and older men.
Correspondence: Alan C. Geller, MPH, RN, Department of Dermatology, Boston University School of Medicine, 720 Harrison Ave, DOB 801A, Boston, MA 02118 (firstname.lastname@example.org).
Accepted for Publication: May 13, 2008.
Author Contributions: All authors had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Geller, Johnson, Miller, and Swetter. Acquisition of data: Johnson, Layton, and Swetter. Analysis and interpretation of data: Geller, Miller, Brooks, and Swetter. Drafting of the manuscript: Geller, Johnson, Brooks, Layton, and Swetter. Critical revision of the manuscript for important intellectual content: Geller, Johnson, and Swetter. Statistical analysis: Miller and Brooks. Obtained funding: Geller and Swetter. Administrative, technical, or material support: Geller, Johnson, and Swetter. Study supervision: Geller and Swetter.
Financial Disclosure: None reported.
Funding/Support: This study was funded by the Harry J. Lloyd Charitable Trust for Melanoma Research.
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: Mitzi C. Rabe, RN, BSN, OCN, was the study coordinator at the University of Michigan. The Melanoma Prevention Working Group provided valuable contributions to the study design and analysis. Frances Scanlon, Scanlon Design, Milton, Massachusetts, designed the “Watch your back” advertisement.