| dc.description.abstract | Objective: The reason for an increase in observed incidence of melanoma of the skin while mortality has increased at a comparatively lower rate is not fully understood, though there are several hypotheses that might explain the phenomenon. We conceptualized a natural history model for melanoma of the skin and explored different hypotheses that might explain the observed changes in incidence and mortality. Methods: We conceptualized and developed an 8-state state-transition model to reflect the natural history and burden of melanoma during the time period of 1992-1996. We used data from the Danish Pathology Database, the Cancer Registry of Norway, biopsy data, and published literature in order to inform calibration targets in the model such as incidence and mortality rates of melanoma, in-situ melanoma incidence rates, and biopsy rates. We used a manual calibration technique to parameterize unobservable parameters, including transition probabilities and detection probabilities. We used the base-case model to explore two hypotheses, “UV exposure” hypothesis and “increased diagnostic activity” hypothesis, to compare model outcomes to observed incidence and mortality rates of melanoma in 2017-2021. Results: Of 81 calibration targets, 60 (74%) were within calibration target “windows.” In order to test the “UV exposure” hypothesis, we needed to increase the probability of developing melanoma in-situ by a factor of 10 in order to match observed local melanoma incidence in 2017-2021. However, after adjusting for 2017-2021 improved survival, observed local stage mortality rates were on average 65% higher than the model’s predicted local mortality rate. In order to test the “increased diagnostic activity” hypothesis, we increased the probability of getting a biopsy in order to match observed incidence and biopsy rates. Even if we increased the probability of detection of a person with local stage undetected melanoma to 0.99, it did not match the observed incidence. We needed to increase the probability of in-situ detection by a factor of 10 and the probability of a healthy person getting a biopsy by 1.25 in order to explain observed biopsy rates. Conclusions: Neither the “UV exposure” theory or the “increase diagnostic activity” theory alone could explain the increase in melanoma incidence and the stabilization in mortality, regardless of whether improved survival rates were used. This would suggest that a combination of the two hypotheses or another explanation would have to explain the observed incidence and mortality trends. However, the limitations of the model structure and parameterization of the model could be improved and explored further. | eng |