Rodríguez MO, Betancourt VM, García RAB, Giance PL

Language: Spanish
References: 0
Page: 1-11
PDF size: 283.59 Kb.

ABSTRACT

Introduction: prediction is a major practice in health care and research nowadays. The development achieved by dental services in Cuba has contributed to reduce the incidence of oral cancer, but it is still a reason for concern that a large number of people are diagnosed with this condition every year. In this context, time series analysis is crucially important for disease surveillance and prediction.
Objective: analyze the behavior of oral cancer incidence and mortality series to generate a prognosis with a three-year prediction horizon in the province of Ciego de Avila.
Methods: a time-series retrospective longitudinal descriptive study was conducted. An analysis was performed of the oral cancer time series, based on an annual incidence frequency for eight years (2007-2014) and a mortality rate for 12 years (2003-2014). Simple, fast, effective methods were used to obtain the prognosis: the least squares method and two-parameter exponential smoothing, which smooths local oscillations making it possible to appreciate the trend in the series. Use was made of the software EVIEWS 4.1.
Results: regarding incidence, a prognosis was made of a point estimation of approximately 73 cases per year, with prediction intervals ranging between 59 and 87 for the next years. Regarding mortality, very narrow prediction intervals were obtained which range between 26 and 50, with a point estimation of approximately 37 deaths per year.
Conclusions: behavior of the oral cancer incidence and mortality series in the province is on the increase, and it is estimated that it will continue to grow in the next three years.