Guidelines for appropriate use of cluster analysis in Public Health

• Grant number: 13/24186-6
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Start date: March 01, 2014
• End date: August 31, 2014
• Field of knowledge: Health Sciences - Collective Health - Public Health
• Principal Investigator: Fernando Frei
• Grantee: Juliana Alves Pegoraro
• Host Institution: Faculdade de Ciências e Letras (FCL-ASSIS). Universidade Estadual Paulista (UNESP). Campus de Assis. Assis , SP, Brazil

Abstract

Pattern recognition is a practice of great importance in healthcare. This practice allows the identification of subgroups of the population who have certain diseases or risk behaviors allowing specific treatments and educational campaigns directed. Cluster analysis is the generic name for a set of methods that can be used to create ratings and consequently identify patterns. Although the methods are simple, there is a wide range to fulfill the same role. In the absence of guidelines, the researchers tend to adopt methods used in newspapers, which largely do not have enough information about the procedures adopted. As with any research, it is necessary to report the methodology clearly so others can replicate the study. Accordingly, this study aims to identify deficiencies in the use of Cluster Analysis - agglomerative hierarchical methods - in health care, and present suitable alternatives for solving these deficiencies, which can serve as guidelines for the appropriate use of Cluster Analysis.