Ten year mortality in a cohort of nonmelanoma skin cancer patients in Denmark.

TO THE EDITOR The incidence of nonmelanoma skin cancer (NMSC) increases worldwide, partly due to the population ageing. Therefore, any impact of NMSC on mortality would have major public health implications. Data on mortality after NMSC are few and remain controversial and the role of co-morbidity has not been clarified. We therefore assessed the association between NMSC in Denmark and mortality throughout a 10-year followup period, and compared with that of a reference cohort from the general population. This nationwide cohort of NMSC was created in 1995. Data were collected from patients’ clinical records collected by all Danish dermatologists, and pathological records collected by the collaborating dermatopathologists. In 1995, Denmark had 5,227,861 inhabitants, and the entire population received tax-supported free health care from the National Health Service. Clinical and pathological records were available for 3,209 patients with an incident diagnosis of NMSC, while 2,759 patients only had pathological records (Table 1). Incident cases of NMSC were determined by record linkage to the Danish Cancer Registry (Storm et al., 1997). This linkage was obtained by the unique 10-digit civil registration number, assigned to all Danish residents by the Civil Registration System, which allows unambiguous record linkage between registries. The Civil Registration System, which has existed since 1968, is updated daily and contains information on vital status, civil status, date of death, and the residence. Through this, a reference cohort of 10 persons for each NMSC patient matched on age, gender, and municipality (31,729 persons) was sampled from the general population. Likewise, mortality, civil status, and migration data for both cohorts were obtained from the Civil Registration System. For the patients, the follow-up period began at the date of diagnosis; for the reference cohort, it began at the date when their corresponding patient was diagnosed (index date) and for both groups it ended at the date of emigration, death, or on 1 November 2004, whichever occurred first. The underlying cause of death was retrieved from the death certificates stored in the National Board of Health (Juel and Helweg-Larsen, 1999). We computed the Charlson Comorbidity Index (Charlson et al., 1987) based on all hospital discharge diagnoses in the National Patient Registry established 1 January 1977, until the diagnosis date of each patient and index date of each reference. As a measure of socio-economic position, we defined three levels of civil status as