You are here

ARC Event: Beyond Proportional Hazards

Wednesday 30 October 2019 09:45 - 18:00

ARC Event: Beyond Proportional Hazards: Statistical methods for assessment of the impact of medical advances and health interventions on longevity and population projections.

This workshop is being delivered by the Actuarial Research Centre’s (ARC) ‘Big Health and Actuarial Data’ Research Programme.


The workshop covers contemporary methods of survival analysis and their use in assessment of possible changes in longevity and population life expectancy due to medical advances.

Technical Content

Cox regression is the most popular method of survival analysis. However it is only applicable under a strong assumption of proportional hazards. This workshop will explore novel advanced methods of survival analysis, which can be used for time-varying hazards and covariates. One of these methods will be the landmark analysis. The workshop will provide a detailed introduction to landmark analysis and a case study on the effect of statins on longevity. Three additional case studies will be further discussed: on stroke, diabetes and the use of hormone replacement therapy (HRT).

Level of detail

The workshop will be sufficient for general understanding of the statistical methods and their application to longevity and population projections.

Target audience

Life and health actuaries; data analysts; actuarial Students.

Prior knowledge

Some knowledge of survival analysis would be beneficial but is not essential.


Attendees will acquire understanding of novel statistical methods and their possible applications to longevity-related tasks in life insurance.

09.45-10.00 Registration

Welcome and Introductions
Andrew Cairns


Cox regression, the proportional hazards assumption and time-varying effects
Ilyas Bakbergenuly

10.50-11.00 Q&A
11.00-11.30 Morning refreshments
11.30-12.10 Linking survival modelling results to longevity improvements
Elena Kulinskaya
12.10-12.30 Q&A
12.30-13.15 Lunch
13.15-13.55 Modelling non-proportional hazards:
time-varying coefficients,   parametric “double Cox” regression and landmark analysis
Elena Kulinskaya
13.55-14.25 Landmark analysis of the use of statins
Ilyas Bakbergenuly
14.25-14.45 Q&A
14.45-15.15 Afternoon refreshments
15.15-15.35 Modelling stroke mortality
Padma Chutoo
15.35-15.40 Q&A
15.40-16.00 Diabetes and its effect on Longevity
Njabulo Ncube
16.00-16.05 Q&A
16.05-16.25 HRT and its effects in morbidity and longevity of women
Nurunnahar Akter
16.25-16.30 Q&A
16.30-16.55 Audience discussion
16.55-17.00 Chair's closing remarks
Andrew Cairns
17.00-18.00 Drinks reception




Staple Inn Hall, High Holborn, London, WC1V 7QJ

Nearest Public Transport

Chancery Lane Station