External Seminar: Leon Danon (Bristol)

[Statistics & Probability Seminar]

Defining the Population Attributable Fraction for transmissible diseases

The population attributable fraction (PAF) is increasingly being used to quantify the contribution of a risk group to an infectious disease burden. Unlike for non-communicable diseases, high risk individuals may increase disease risk for the wider population in addition to themselves, therefore alternative methods are required to estimate the PAF for infectious diseases. We use a mathematical modelling approach to compare existing methods of calculating the PAF and develop a new method that addresses shortcomings. We demonstrate that current methods result in varied PAF estimates and derive two simple formulae that can be used to estimate PAF for emerging transmissible diseases and for transmissible diseases close to endemic equilibrium. Current methods include chains of transmission which are not dependent on the risk group, and result in larger PAF estimates than traditional methods. Policy makers and users of PAF statistics should be aware of the differences between methods.