Share this Article
Ghadimi MR, Mahmoodi M, Mohammad K, Rasouli M, Zeraati H, Fotouhi A
Correspondence: Mahmood Mahmoodi, email@example.com
Introduction Oesophageal cancer is one of the most common causes of cancer mortality in developing countries, including Iran. This study aimed to assess factors affecting survival of patients with oesophageal cancer using parametric analysis with frailty models.
Methods Data on 359 patients with oesophageal cancer was collected from the Babol Cancer Registry for the period 1990–1991. By 2006, the patients had been followed up for a period of 15 years. Hazard ratio was used to interpret the risk of death. To explore factors affecting the survival of patients, log-normal and log-logistic models with frailty were examined. The Akaike Information Criterion (AIC) was used for selecting the best model(s). Cox regression was not suitable for this patient group, as the proportionality assumption of the Cox model was not satisfied by our data (p = 0.007).
Results Multivariate analysis according to parametric models showed that family history of cancer might increase the risk of death from cancer significantly. Based on AIC scores, the log-logistic model with inverse Gaussian frailty seemed more appropriate for our data set, and we propose that the model might prove to be a useful statistical model for the survival analysis of patients with oesophageal cancer. The results suggested that gender and family history of cancer were significant predictors of death from cancer.
Conclusion Early preventative care for patients with a family history of cancer may be important to decrease the risk of death in patients with oesophageal cancer. Male gender may be associated with a lower risk of death.
Keywords: AIC, inverse Gaussian frailty, oesophageal cancer, survival analysis
Singapore Med J 2012; 53(5): 336–343