Analysis and modelling of emerging infections: an interview with Ilaria Dorigatti

Interview by Francesca Arici and Anna Maria Cherubini

  • Born inItaly
  • Studied inItaly
  • Lives inUnited Kingdom


Ilaria Dorigatti is Lecturer at Imperial College London and takes part in the COVID-19 Response Team of the College. Previously,  she has been a member of the Ebola Response Team and has contributed to the characterisation of  Zika virus.

Can you describe your work as a mathematician at the MRC Centre for Global Infectious Disease Analysis? 

My research focuses on the analysis and modelling of emerging infections, including respiratory viruses such as SARS-CoV-2, the causative agent of COVID-19, and vector borne diseases such as dengue, Zika and Yellow Fever. It aims to address public health priorities and answer real-world questions on disease transmission and control strategies. My work is based on epidemiological data and consists of the development of mathematical and statistical models to better understand the mechanisms, drivers, and heterogeneities of transmission and disease. From a practical perspective, the work of a mathematical modeller consists of developing models that allow inferences to be made and test hypotheses using observed data, rigorous statistical methods and an evidence-based approach. The MRC Centre for Global Infectious Disease Analysis is a very dynamic place. Beyond having the privilege of working with world-leading experts in the field, I have regular opportunities to interact with colleagues who have different backgrounds from mine, e.g. in statistics, medicine, biology, computer science and ecology, which is incredibly stimulating and extremely helpful to piece together a full picture of the disease transmission puzzles observed in the data.

My research addresses public health priorities and answers real-world questions on disease transmission and control strategies.

How did you start working on epidemiology? 

Since my studies at the Department of Mathematics of the University of Trento, Italy, I have always been fascinated by the applications of maths to answer real world questions, particularly to study the dynamics of infectious diseases in animal and human populations. My very first research project was about modelling the spatial dynamics of the highly pathogenic H7N1 avian influenza virus in Italy. During my PhD, I had the opportunity to visit Imperial College and work with Neil Ferguson and Simon Cauchemez at the MRC Centre for Global Infectious Disease Analysis (then the MRC Centre for Outbreak Analysis and Modelling).  This was a fantastic experience and I loved the centre’s research environment. In 2011, I moved to London to take a fixed-term postdoc position to investigate the drivers of the third wave of infection of the 2009 H1N1 virus in England. In 2015 I was awarded an Imperial College Research fellowship and in 2018 I received a Sir Henry Dale fellowship, which was then followed by a lectureship position in the School of Public Health within the Faculty of Medicine, Imperial College London.

You studied SARS-COV2 transmissibility by looking at the spread of coronavirus in the village of Vo’ Euganeo in Italy. What is, in your opinion, the most striking insight this research has to offer?

Our study published in Nature offers several striking insights. Perhaps the most important is about the large proportion (mean of 42.5%) of asymptomatic infections (i.e. not showing symptoms at the time of swab testing and not developing symptoms afterwards) detected in two surveys that were conducted in Vo, Italy, soon after the first COVID-19 death was detected in Italy. This finding, along with the evidence of asymptomatic and pre-symptomatic transmission, highlighted very early on in the pandemic the challenges associated with the control of COVID-19 and the need to strengthen disease surveillance. The experience of Vo also serves as an example of epidemic suppression, which was possible due to the timely implementation of interventions in the early phases of the epidemic.

Which is your current research?

I am currently analysing serological data and investigating heterogeneities in SARS-CoV-2 transmission, as well as the potential impact of seasonality across spatial scales. My arbovirus work currently focuses on better understanding of spatiotemporal dynamics and the potential impact of novel interventions as well as characterising within-host viral dynamics.

In your opinion, what can policymakers learn from these studies? And how much impact do scientific analyses have on health policies in general and in the COVID-19 case in particular?

In my experience, scientific studies have generally had a positive impact on informing policy decisions. As a WHO Collaborating Centre, we provide support to the WHO as well as many other public health agencies, not only during outbreaks and in pandemic times. Our models and analyses, along with those of other experts in the field, provide key insights into situational awareness and the potential impact of interventions, thus informing decision making, response planning and the development of guidelines and recommendations.

In my experience, scientific studies have generally had a positive impact on informing policy decisions.

During the COVID-19 pandemic, I think that key scientific results of public health relevance have generally reach decision makers. Different governments use different pathways for gathering scientific evidence. In the UK for instance, the government gathers scientific advice through multiple groups – these may have different perspectives and use different approaches, which are important to take into consideration to assess the consistency and uncertainty of each single model. Clearly, data sharing – which can be confidential when dealing with sensitive data in uncertain times – is an important aspect of this process.

That said, there is an important and clear distinction to be made between science and policy making. While mathematical modelling provides useful insights and future scenarios, policy decisions are taken based on considerations that typically go beyond modelling results.

There is an important difference between science and policy making. While mathematical modelling provides insights and future scenarios, policy decisions are based on considerations that typically go beyond modelling results.

Could you tell us about the particular challenges related to COVID-19 in comparison to other diseases you have been working on, like Zika or Ebola?

These three diseases have different mechanisms of transmission and very different epidemiological characteristics. COVID-19 is a respiratory disease, which means that it is passed from one person to another via droplets exhaled while breathing, coughing, or sneezing. Zika instead is a vector borne disease which is transmitted from person to person through the bites of an infected mosquito. Ebola is transmitted via direct contacts with an infected person or deceased patients.

While SARS-CoV-2 and Zika have similar basic reproduction numbers – a key measure of the contagiousness of the virus, which is related to the herd immunity threshold – Zika relies on the presence of vectors for its transmission and geographical spread. SARS-CoV-2 spread rapidly across the globe for its respiratory nature and the challenges associated with mild or asymptomatic infection and transmission. Ebola instead is transmitted upon the occurrence of symptoms and is considerably less transmissible than SARS-CoV-2, these epidemiological characteristics contributed to the control efforts that avoided global spread.

The probability to die of COVID-19 is currently estimated to be around 1% overall, with a strong age-gradient and a much higher risk for the elderly and vulnerable populations. The probability to die of Ebola is much higher in comparison (around 70%). But the speed and scale of the COVID-19 pandemic is such that the number of infections requiring hospitalisation and critical care can rapidly exceed hospital capacity – this is one of the global public health challenges that we are facing since the emergence of the virus. There are also current challenges in the scaling up of testing and contact tracing as well as communication challenges and adherence to public health recommendations.

Can you tell us something about your experience as a member of the WHO Ebola Response Team?

Being part of the WHO Ebola Response Team during the 2013-2016 epidemic in West Africa was an incredible experience for me, because it was the first time that I had the opportunity to be part of a team working on real-time outbreak analysis and I learnt so much! I learnt not just about data limitations, methods, modelling and teamwork but also about communicating sensitive results and delivering under time constrains. Our work informed WHO’s situational awareness, planning and response – these are core values for my research and for the MRC Centre for Global Infectious Disease Analysis.

In your opinion, what contributions can modelling bring to an uncertain world like the one we live in? and what is to be learned from the ordeal we have been (and still are) going through?

Unfortunately modelling cannot predict the future with certainty. However, it can provide scenarios and estimates of the uncertainty. In this sense, modelling is extremely useful and can be used to project forward in time and explore ‘what if’ scenarios. Modelling also allows to explore counterfactuals, i.e. look backwards and quantify the potential impact of interventions had they been implemented differently. For instance, it has been demonstrated that the timing of interventions relative to the stage of the epidemic has a critical impact on the outcome of an epidemic. Countries that put in place clear strategies to reduce transmission early on, in the exponential phase of the epidemic had a better outcome, i.e. lower infection and mortality rates.

In my opinion, this pandemic has demonstrated the importance of test & trace strategies and of scaling up surveillance and preparedness planning. In the absence of a vaccine, contact tracing in its various forms (standard, i.e. forward-looking as well as backward contact tracing to identify hot spots of transmission) and non-pharmaceutical interventions are the only strategies available for disease control. Communicating clearly and consistently with the public is essential to promote adherence to the guidelines.

Communicating clearly and consistently with the public is essential to promote adherence to the guidelines.

How have the pandemic and the lockdown impacted your research?

For the nature of my work, I could conduct my research remotely, so the lockdown has not stopped or significantly changed the way I work. The main exception is teaching which continues to be delivered remotely at least until next year. As it always happens in outbreak situations, the workload increases, especially in the early phases of a pandemic and there is some pressure to deliver estimates and results in a timely fashion. With the support of the team, we all kept going.

On the plus side, we could enjoy lunch all together for weeks in a row, something that had never happened before!

And how has it impacted on the rest of your work and life? 

As for everyone else, the lockdown has impacted my day-to-day and social life. I miss seeing my students and colleagues and having in-person rather than virtual catch-ups. During the lockdown, we organised work in shifts with my husband – beyond working during ‘out of office’ hours – to make sure that we would get the work done and the same time look after the children and do homeschooling. On the plus side, we could enjoy lunch (however quick and simple) all together for weeks in a row, something that had never happened before!