Your Cart

Your cart is empty

Home  >  Transasctions of NAMP VOL2

17. Logistic Regression for Cancer Data BY G.O. Korter and J.O.Omolehin Transactions of NAMP Volume2,(November, 2016), pp 163 – 170
Sale price: $5.00


More than 60% of the world’s total new annual cancer cases occur in Africa, Asia and Central and South America. The aim of this study was to build a predictive model for the two possible outcomes of cancer patients and to examine which of the several types of cancer was more deadly. Secondary data of 335 patients aged 11 to 90 years who received treatment for liver, lung, colon, colorectal, prostate, breast or skin cancer at LAUTECH teaching hospital between 2004 and 2010 was used for this analysis. Logistic regression analysis was conducted. The Hosmer and Lemeshow and Likelihood Ratio tests were used to determine the fit and significance of parameters of the model. Only the type of cancer suffered by patients contributed significantly to the prediction model. The odds of dying for patients with lung cancer were about 4 times that of other types of cancer. However, the incidence of liver, lung, colon, colorectal, prostate, breast and skin cancer was prevalent across patients aged 11 and 90 years, irrespective of sex. Lung cancer was found to be more deadly than other types of cancer observed in the sample.

Keywords: Cancer, deaths, developing countries, logistic regression, prevention