Prediction of cancer incidence in the Nordic countries: empirical comparison of different approaches

Abstract

Prediction of the future number of cancer cases is of great interest to society. The classical approach is to use the age-period-cohort model for making cancer incidence predictions. We made an empirical comparison of different versions of this model, using data from cancer registries in the Nordic countries for the period 1958-1997. We have applied 15 different methods to 20 sites for each sex in Denmark, Finland, Norway and Sweden. Median absolute value of the relative difference between observed and predicted numbers of cases for these 160 combinations of site, sex and country was calculated. The medians varied between 10.4 per cent and 15.3 per cent in predictions 10 years ahead, and between 15.1 per cent and 32.0 per cent for 20 year predictions. We have four main conclusions: (i) projecting current trends worked better than assuming that future rates are equal to present rates; (ii) the method based on the multiplicative APC model often overestimated the number of cancer cases due to its exponential growth over time, but using a power function to level off this growth improved the predictions; (iii) projecting only half of the trend after the first 10 years also gave better long-term predictions; (iv) methods that emphasize trends in the last decade seem to perform better than those that include earlier time trends.

The future burden of cancer in England: incidence and numbers of new patients in 2020

“…For this analysis of the future cancer rates and numbers of cancer cases in England, we have used a method of estimation that was developed in a comprehensive and systematic analysis of cancer trends in the Nordic countries (Møller et al, 2002. Møller and co-workers used the long data series in the Nordic countries to develop a large number of predictions of present rates as would have been forecast 20 years ago, and compared the predicted rates with those actually observed, identifying a set of analysis options that tended to give the most accurate predictions.…”

“…A recent approach to projecting incidence rates (Møller et al, 2002) was adapted for most of the cancer sites, based on a standard age-period-cohort model (Osmond, 1985), but with some modifications which have been shown empirically to improve the predictions . The multiplicative relationship between incidence and the covariates in the standard model produces predictions in which the rates change exponentially with time.…”

“…We considered the changes in these parameters, both in absolute and in relative terms. The percentage change in the annual number of cases from 2001 to 2020 was divided into one part due to increased risk of being diagnosed with cancer, and another due to changes in the population size and age distribution (Møller et al, 2002). The prediction package NORDPRED written in R is available online (Nordpred, 2006).…”

“…Secondly, from the point of view of cancer care and cancer services provision, the burden of cancer is more usefully measured as the total number of persons with cancer who require diagnostic, therapeutic, supportive or palliative services (Bray and Møller, 2006). Several large studies have forecast future cancer rates and numbers, using a variety of statistical methods (Møller et al, 2002;New Zealand Ministry of Health, 2002;Scottish Executive, 2004;Clements et al, 2005). The most common methods relate incidence to the three interdependent time dimensions of age, calendar period and generation.…”

Modelling the impact of population-based cytologic screening on cervical cancer incidence and mortality in Hong Kong: an age–period–cohort approach

“…To examine and quantify the potential public health impact of this new preventive programme relative to the status quo of haphazard screening, we adopted the age -period -cohort (APC) approach, which has been used extensively to predict future incidence and mortality trends under different public health intervention scenarios (Osmond, 1985;Dyba and Hakulinen, 2000;Møller et al, 2003;Quinn et al, 2003). Our primary objective was to project the number of incident cancers potentially averted and years of life saved (YLS) attributable to the new organised screening programme through 2016, by modelling historical incidence and mortality data from 1972 to 2001 using maximum likelihood and Bayesian methods.…”