TY - JOUR TI - Gender, age, and screening differences in individuals diagnosed with malignant melanoma in the United States DO - https://doi.org/doi:10.7282/T3R78D0B PY - 2012 AB - Melanoma is the rarest but most lethal form of skin cancer. This dissertation focuses on three salient issues in melanoma research –pediatric incidence, completeness of ascertainment in cancer registries, and the role of screening. Chapter 1 addresses melanoma incidence in children in whom 40-60% of melanoma cases may be initially misdiagnosed. The purpose was to examine differences between children/adolescents and adults in demographics and clinical characteristics of melanoma. Cases diagnosed from 1995-2008 were identified using the Cancer in North America (CINA) Deluxe database from the North American Association of Central Cancer Registries (NAACCR). Frequency distributions and incidence rates were tested for differences using chi square statistics, rate ratios, and annual percent change. Results show that children were diagnosed at later stages (x2 = 63.59; p<.0001) and were more likely to have thicker lesions (x2 = 22.3; p<0.0001) than adults. Questions are raised about the role of hormonal/reproductive factors contributing to age and gender differences. Because of the growing evidence for under-reporting of melanoma incidence we investigated this issue in New Jersey (Chapter 2). We surveyed dermatologists to identify why melanoma may be underestimated and to quantify the extent of reporting delay. We also estimated the missing melanoma cases using a capture-recapture analysis. Using log-linear models we approximated that 817 melanoma cases were missed annually, most likely from physicians and pathology labs. These estimates can be used to improve the accuracy of melanoma incidence rates and to make targeted adjustments for reporting. In Chapter 3 we examined long term melanoma survival rates for skin self-examiners (SSE), which is a useful and inexpensive screening method that has the potential to reduce the risk of advanced disease. Cases were diagnosed in 1987-1989, followed through 2007, and analyzed using competing risks (CR) analysis. Cumulative incidence functions and proportional hazards regression models were fitted. The cumulative incidence curves by SSE were not statistically different (p=0.32) for death due to melanoma in the presence of CR. Skin awareness (HR= 0.49, p=0.002) was associated negatively and thickness (HR= 1.21, p<0.001) was associated positively with melanoma death. Although we did not find a significant association between melanoma mortality and SSE, we have confirmed previous findings of a protective association with skin awareness. KW - Public Health KW - Melanoma--United States KW - Medical screening--United States KW - Children--Health and hygiene LA - eng ER -