Predictions |
These options allow computation of short and long term predictions of cancer incidence and mortality in the Nordic countries. Predictions are uncertain, and the results should always be interpreted with caution. The authors do not accept any responsibility or liability in regard of the reliance on, and/or use of the results. The forecast populations stem from the national statistical bureaus. |
Short term predictions
(up to five years)
The short term predictions are computed using a program developed at the
IARC, based on Tadeusz Dyba and Timo Hakulinen methods. For a selected combination of data type, country, cancer, sex and a 'latest year' of observation, the predictions are performed using age-period models which require at least six consecutive years of data with at least 50 cancer cases or cancer deaths (all ages) recorded per year. The results are presented by age groups and for all ages combined, together with their corresponding prediction interval. Graphical representation of the results is provided. |
Long term predictions
(more than five years)
Long term predictions are computed using the NORDPRED package developed by
Harald Fekjær and Bjørn Møller at the
Cancer Registry of Norway. For a selected combination of data type, country, cancer, sex and a 'latest year' of observation, the predictions are performed using an age-period-cohort model which requires at least fifteen years of consecutive data (in order to build at least three 5-year periods), with at least 100 cancer cases or deaths (all ages) recorded per 5-year period. It can predict up to four 5-year periods. The results are presented by age groups and for all ages combined. For each predicted period, the change in the number of predicted cancer cases or deaths to the latest period of observation is divided into one part due to change in risk of getting or dying from the cancer, and another one due to changes in the population size and age distribution. Graphical representation of the results is provided. |
This work is a collaboration between:
1. Jacques Ferlay at IARC 2. The NORDCAN secretariat 3. and the following experts: Bjørn Møller, The Cancer Registry of Norway Tadeusz Dyba Timo Hakulinen |
References: 1. Bray F, Møller B. Predicting the future burden of cancer. Nat Rev cancer. 2006;6:63-74 2. Dyba T, Hakulinen T, Päivärinta L. A simple non-linear model in incidence prediction. Statistics in Medicine 16: 2297-309, 1997. 3. Dyba T, Hakulinen T. Comparison of different approaches to incidence prediction based on simple interpolation techniques. Stat Med. 2000 Jul 15;19(13):1741-52. 4. Hakulinen T, Dyba T. Precision of incidence predictions based on Poisson distributed observations. Statistics in Medicine 13: 1513--23, 1994. 5. Møller B, Fekjaer H, Hakulinen T, Sigvaldason H, Storm HH, Talback M, Haldorsen T (2003) Prediction of cancer incidence in the Nordic countries: empirical comparison of different approaches. Stat Med 22(17): 2751–2766 6. NORDPRED software package available at: http://www.kreftregisteret.no/en/Research/Projects/Nordpred/ (accessed on 17/12/2008) 7. Stata macros for short-term predictions available at: http://www.encr.com.fr/stata-macros.htm 8. United Nations, World Population Prospects, the 2008 revision http://www.un.org/esa/population/unpop.htm |
NORDCAN © 2009 Association of the Nordic Cancer Registries - All Rights Reserved. |