A Maltese mathematician mapped out the way swine flu spread across the island in 2009 and 2010 and came up with a set of models that can help predict the outbreak of future epidemics.

The models, based on four data sets, allow for an early warning of an oncoming epidemic and for the prediction of the rate at which it will spread, Vincent Marmará, who is also a statistician, said.

This, he added, could help ensure timely action is taken and the right amount of
vaccines bought.

Together with researchers Alex Cook and Adam Kleczkowski, Mr Marmará outlined the findings in a paper entitled ‘Estimation of force of infection based on different epidemiological proxies: 2009/2010 influenza epidemic in Malta’. In the paper, published in the journal Epidemics, and forming part of his PhD in mathematics (epidemiology), he focuses on prediction and control of epidemic outbreaks based on limited information.

The researchers resorted to data about influenza collected by the Malta Health Promotion Department during the swine flu (H1N1) epidemic when the first cases emerged in Malta in 2009. The researchers focused on data collected by eight physicians selected by the department between September 2009 and June 2010. During that period, 52,016 patients consulted the eight doctors and 4,544 were diagnosed with common influenza.

The amount of patients swabbed for swine flu amounted to 1,847 and those who tested positive reached 622. Five patients died.

Between January and February 2010, vaccine became available to everyone, so March of that year was considered to mark the end of the epidemic.

Equipped with such information the researchers drew up four data sets based on consultations, patients diagnosed (with the common flu), patients swabbed and positive cases (for swine flu).

They also took into consideration several factors including under-reporting.

Using established and up-to-date scientific modelling techniques, the four data sets were used to create four models, which could be applied to predict epidemic patterns in Malta. These models can be tailored to particular diseases by inserting certain parameters such as information about the duration of an illness.

“The innovative part of the research is that it allows for the use of the four different data sets to confirm results and predict the actual outbreak,” Mr Marmará said.

So, for example, if a case of Ebola is detected in Malta, the amount of consultations – together with information about the disease drawn from the experience of other countries – can help predict the rate at which it will spread.

This can then be crosschecked mathematically with data about patients diagnosed with symptoms.

Download Paper: Marmara et al (2014)