This five-year research programme was commissioned in 2016 to develop new methods for assessing basis risk and evaluating longevity improvement based on Big Health and Actuarial Data.

Programme objectives

  • identification and quantification of the key factors affecting mortality/longevity such as lifestyle choices, medical conditions, and/or interventions
  • modelling of temporal changes in the factors affecting morbidity and mortality
  • evaluation of plausible scenarios in mortality trends due to particular medical advances or lifestyle changes in the population of insureds of relevance to the actuarial community
  • tools to forecast longevity risk of a book based on realistic scenarios of uptake of various health behaviours and/or interventions, or of particular disruptions to population health.
The multidisciplinary University of East Anglia (UEA)/Aviva team was led by Prof Elena Kulinskaya and included actuaries, statisticians, health economists, computer scientists, and medical researchers. The funding supported three PhD students and a postdoctoral researcher.  Below is a summary of what was achieved to date.

Research outputs

Life expectancy calculator

On 3 September 2020, the research programme launched mylongevity.org to help illustrate the effect of various medical and lifestyle factors on life expectancy. The Mylongevity app uses big data from anonymised electronic health records in England and Wales to calculate life expectancy, taking into account socio-demographic and health characteristics such as chronic diseases.
 
The new app could potentially bring practical benefits, such as helping people plan for retirement. It could also contribute to scientific understanding of how certain behaviours and lifestyles affect longevity.
 
The app uses calculations based on UK life expectancy figures published by the Office for National Statistics and then refined by the researchers. The ARC Programme identified and quantified the key factors affecting mortality and longevity and translated the results of their analysis into life expectancies for different risk profiles.
 
The R package behind MyLongevity.org is available on GitHub for companies to download (Software Rtools40 will need to be installed). It can be used as an app in bulk format, e.g. 100,000 clients, or adapted using company parameters, population or weights. The R package currently works under R version 3.4.2 (64-bit) only. The package includes programs to calculate life expectancy based on the Cox proportional hazards model and landmark analysis combined with the Gompertz baseline risk, as used in the programme’s MyLongevity app. The R package also provides co-called double-Cox modelling for stroke, TIA, HRT and type 2 Diabetes.

Publications

Presentations

 
The research team have presented their findings at top actuarial and statistical conferences, including the Actuarial Teachers and Researchers Conference (Dublin 2021), International Conference in Actuarial Science, Data Science and Finance (2020), Insurance Data Science Conference (Zurich 2019), International Congress of Actuaries (Berlin 2018), International Biometric Society Conference (Barcelona 2018), International Society for Clinical Biostatistics Conferences (Vigo 2017, Belgium 2019), Mortality and Longevity Symposiums (2016, 2019), Life Conference (2017), Royal Statistical Society conference (2017, 2019) and Actuarial Research Conference (2016).

Events calendar

No results found.