ASSOCIATION RULES OF SOCIAL DETERMINANTS AND DEATH FROM ORAL CANCER IN BRAZIL: A DATA MINING STUDY

• Published in: Oral surgery, oral medicine, oral pathology and oral radiology Vol. 134; no. 3; p. e203
• Main Authors: Pires, Alessandra Laís Pinho Valente, De Matos Amorim, Marília, Lopes, Murilo Cruz, Da Silva, Adriana Mendonça, De Oliveira, Ana Carla Barbosa, Calumby, Rodrigo Tripodi, Freitas, Valéria Souza
• Format: Journal Article
• Language: English
• Published: Elsevier Inc 01-09-2022
• ISSN: 2212-4403; 2212-4411
• DOI: 10.1016/j.oooo.2022.01.627

Determining association rules related to social determinants and death from oral cancer in Brazil.
Data mining study from the Information System of Hospital Registry of Cancer for individuals with oral cancer from 1997 to 2017. Multiple preprocessing techniques and standardization of the data were applied, in addition to to attribute engineering methods. Subsequently, the FP-Growth algorithm was applied with selection of rules with minimum acceptable support and confidence, in addition to analysis by specialists in the application area.
A total of 72,039 individuals were studied. The majority of the population was male, elderly, White, with less education, and drinkers and smokers. Most of the lesions were located on the tongue, diagnosed in advanced stages, and treated with radiotherapy, and 41.6% of the population died. The data mining analysis indicated an association of death with low level of education, absence of race/color information, male sex, and occurrence in hospitals in the southeast.
The results of this study can assist in decision making and might contribute to the estimation of the risk of death in the population and as a basis for the definition of public policies.