[Study of variables associated with skin cancer in Chile using principal component analysis]

Actas Dermosifiliogr. 2006 May;97(4):241-6. doi: 10.1016/s0001-7310(06)73391-9.
[Article in Spanish]

Abstract

Background: The incidence of skin cancer in Chile has increased in recent years.

Objective: To associate variables with skin cancer in Chile through indices generated using multivariate descriptive statistical techniques.

Material and method: During May 2004, information was gathered from demographic, meteorological and clinical data from Chile corresponding to fiscal year 2001, the latest complete, official information available for the country's Health Services as a whole. The variables developed by the following were studied: the National Statistics Institute (INE), the Ministry of Health (MINSAL), the Ministry of Planning and Cooperation (MIDEPLAN), the National Health Fund (FONASA), the Chilean Meteorological Directorate, Federico Santa María Technical University and the Directorate-General for Water. A Principal Component Analysis (PCA) was then performed on the data obtained.

Results: The first three principal components were selected, with a cumulative explained variance percentage of 54.48 %. The first principal component explains 24.92 % of the variance, and is related to climatic and geographic variables. The second principal component explains 15.77 % of the variance, and is mainly related to FONASA's beneficiary population and the poverty rate. The mortality rate from skin cancer runs significantly against this component. The third principal component explains 13.79 % of the variance, and is related to population characteristics, such as total catchment population, female population and urban population.

Conclusion: Performing PCA is useful in studying the factors associated with skin cancer.

MeSH terms

  • Adult
  • Aged
  • Catchment Area, Health
  • Chile / epidemiology
  • Dermatology
  • Female
  • Geography
  • Hospitals / statistics & numerical data
  • Humans
  • Male
  • Medical Indigency / statistics & numerical data
  • Meteorological Concepts
  • Middle Aged
  • Neoplasms, Radiation-Induced / epidemiology
  • Poverty / statistics & numerical data
  • Principal Component Analysis
  • Risk Factors
  • Rural Population / statistics & numerical data
  • Skin Neoplasms / epidemiology*
  • Sunlight / adverse effects
  • Urban Population / statistics & numerical data
  • Workforce