Future modelling of chronic diseases: foresight and beyond

Abstract

Listen

Translate article

BackgroundMany chronic diseases are interrelated and their effects under changing exposures need to be better understood. Policy makers and planners need to understand what the current distributions of avoidable chronic disease are, among whom, and how they are likely to develop in the future, particularly their effect on different populations, what will be the health consequences of the extrapolated trends, and how much these consequences can be attenuated with what we currently know and might come to know. Modelling the effects by evidence-based extrapolation, incorporating and attributing the epidemiology of related diseases, can give rise to straightforward estimates of incidence and death rates for the most common related conditions for the next 50 years. This modelling can be done by making basic assumptions about plausible change of risk in disease rates tracking into adulthood using established likelihoods from current trends. These rates in turn are used to compare predicted illness and mortality rates with those that arise from demographic extrapolation from existing current mortality rates. These figures can be used to revise estimates of the healthy working population, by removing death rates and accounting for sickness in a manner that incorporates known and current changes in disease incidence. The model builds on work originally developed for Foresight Tackling Obesities (UK) and subsequent work modelling obesity and related diseases in a further 36 countries. MethodsA microsimulation model was used to project future health of each of possibly millions of individuals with a given demography (of any region) through to a given year, and various scenarios were simulated. Competing risks were examined in real simulated time. Related diseases and associated health-care costs were calculated on the basis of trends in risk factors distributed among these individuals. In the case of obesity, 13 diseases were considered. We simulated three hypothetical future scenarios: no reduction and 1% and 5% reductions in body-mass index (BMI). Ultimately, health and other costs incurred or saved can be compared with the costs of intervention. The simulation model was developed in discrete modules to enable radical change and updating of assumptions and parameters. We did not apply any future discounting for this project. FindingsSmall reductions in risk factors can have substantial effects on future burdens of disease and avoidable health-care costs. In the UK, 1% reduction in BMI rates will save £15·5 billion, whereas in the USA the medical costs will be reduced by US$686 billion. With a 5% reduction in BMI, medical cost savings in the UK will be £61·8 billion and in the USA $1·93 trillion. The figures are substantial for other countries too, reaching $45 million for Russia, $1·8 billion for Mexico, and $4·8 billion for Brazil. All savings will be achieved by 2050. InterpretationThis predictive modelling has significant resonance with policy makers. Using sensitivity analysis we can test the outcome of interventions at a national or subnational level over a timescale that is difficult to measure by conventional evaluative methods. FundingGlaxoSmithKline (grant number 27875780), Robert Wood Johnson Foundation (numbers 260639 and 61468 and 66284), Centers for Disease Control and Prevention (U48/DP00064-00S1 and 1U48DP001946).