L Päivärinta | PubMed Health

PubMed1997

A simple non-linear model in incidence prediction.

A simple model is proposed for incidence prediction. The model is non-linear in parameters but linear in time, following models in environmental cancer epidemiology. Assuming a Poisson distribution for the age and period specific numbers of incident cases approximate confidence and prediction intervals are calculated. The major advantage of this model over current models is that age-specific predictions can be made with greater accuracy. The model also preserves in the period of prediction the age pattern of incidence rates existing in the data. It may be fitted with any package which includes an iteratively reweighted least squares algorithm, for example GLIM. Cancer incidence predictions for the Stockholm-Gotland Oncological Region in Sweden are presented as an example.

Adult↗

PubMed1987

Testing equality of relative survival patterns based on aggregated data.

The relative survival rate is defined as the ratio of the survival rate observed in a patient group under consideration to the survival rate expected in a group of people similar to the patient group at the beginning of the follow-up interval, with respect to all possible factors (e.g., age and sex) affecting survival, except the disease under study. Survival from cancer and other chronic diseases is often measured by this quantity, which is adjusted for the effect of mortality attributable to competing risks of death. In this paper, maximum likelihood ratio tests are constructed on the basis of aggregated data for testing the equality of relative survival rates between patient groups against proportional hazards and general alternative hypotheses. The tests are applied to the Finnish nationwide data on colon cancer patients with nonlocalized tumors as reported to the Finnish Cancer Registry. Simulation studies show that the maximum likelihood ratio tests compare favorably with alternative methods proposed earlier. Moreover, the maximum likelihood ratio tests are more extensive in coverage and are based on more applicable alternative hypotheses than the other test statistics. Finally, an extension to proportional hazards regression models of the relative survival rates is suggested.

Biometry↗