Estimating the population effectiveness of the national seasonal influenza vaccination programme in Scotland
Stephen Corson (University of Strathclyde)
Friday 6th February, 2015 15:00-16:00 Maths 516
Every winter, for about 8-10 weeks, the United Kingdom experiences a seasonal influenza epidemic that affects the morbidity and mortality of thousands of its citizens and puts increased pressure on NHS primary and critical care services. To limit the health care burden associated with this disease the NHS vaccinates those who are at highest risk of influenza related morbidity and mortality. Historically, members of the population aged 65 and over only received the vaccine if they belonged to a particular risk group. In 2000 there was a change in policy; all persons aged 65 and over are now eligible for the vaccine regardless of risk group. While there is evidence to suggest that the influenza vaccine is effective at preventing influenza like illness and its complications, the population effectiveness of the vaccination programme is not fully understood, particularly in Scotland.
In this presentation we will use General Additive Modelling (GAM) techniques to derive a Poisson regression model to evaluate the population effectiveness of the national seasonal influenza vaccination programme in Scotland. We shall discuss how we fit the model to routine influenza surveillance data collected in Scotland and how the model's predictions can be used to determine if the outcome of interest (e.g. deaths, influenza related hospitalisations) is higher than expected after the change in policy. We will also quantify how many deaths and emergency hospitalisations this change in policy may have prevented.