Box-Jenkins Method of analysing rate of cancer deaths at a public university hospital in east coast of Malaysia

摘要

Cancer is one of the principal causes of death in Malaysia. This study was performed to determine the pattern of rate of cancer mortality at a public hospital in Malaysia over an 11 year period from year 2001 to 2011; the best fitted model using Box-Jenkins Method of analysis to forecast the rate of cancer mortality and to forecast the rate for the next two years (2012 to 2013). The Secondary data for this study was obtained from a public university hospital located at the East Coast of Malaysia. The data were medical records of patients who died due to cancer from the year 2001 until 2011, with 663 cases. The cancer cases were classified according to ICD10 (10^th Revision International Classification of Diseases). The medical records include socio-demographic background of patients such as registration, age, gender, race, ward and diagnosis. SPPS Version 19 and Minitab 16.0 were used to analyse the data. Malay patients (88.10%) and males (51.30%) were the highest percentages for the overall 11 years death rate of cancer at this hospital. Lung and breast cancer were the most cancer deaths among both gender. The highest death rate (29.60%) is also among patients aged 61 and above. The best model used for forecasting the rate of cancer deaths is ARMA (1, 3). The forecasting values for the rate of cancer deaths for January 2012 to April 2012 range from 5.91% to 6.13%. However, the forecast saw a stable overall cancer mortality trend of 6.14% for May 2012 to November 2013.