Population density and mobility affect the spread of infectious disease.

CDR data can support us to better model the spread of infectious disease by improving our understanding of population distributions and the connectivity between different areas.

By better understanding where the disease is likely to spread next, we can prioritise the implementation of controls to stop or slow the spread. These types of insights can be especially useful in low- and middle-income countries (LMICs) where resources may be more limited and there is therefore more need to target interventions effectively.

Flowminder demonstrated the effectiveness of CDR data to model the spread of cholera in Haiti during the 2010 outbreak. Importantly, the use of CDR data allows the spread of the disease to be modelled and predicted as an epidemic progresses, whereas some other types of data can only be used retrospectively, meaning that it could be used to forecast the spread of disease and inform interventions.

Similarly, CDR data has been used to predict the spread of dengue in Pakistan. Weslowski et al. were able to model the 2013 dengue epidemic in Pakistan by combining CDR and climate data. By combining maps of climatic suitability and connectivity between areas the researchers were able to generate fine-scale, dynamic risk maps with direct applications to dengue control and epidemic preparedness.

A partnership between the International Telecommunication Union, the University of Tokyo, and telecommunications regulators and MNOs from Guinea, Liberia and Sierra Leone has also shown the value of CDR data in preparing for and responding to epidemics, with particular focus on Ebola virus. Their analysis of population mobility in Sierra Leone, for example, highlighted the importance of controlling the epidemic near borders and not just in central areas. Researchers from Flowminder and Harvard School of Public Health have also highlighted the role CDR data could play in responding to Ebola epidemics in Sierra Leone and the importance of protocols for the rapid sharing of CDR data in response to public health emergencies.

The effectiveness of CDR data in predicting the spread of infectious disease has also been demonstrated in high-income countries such as Norway and Singapore. Telenor, a Norwegian MNO, partnered with the Norwegian National Institute of Health, University of Oslo, and Norwegian Computing Centre to model the spread of SARS-CoV-2 virus in Norway. In Singapore, CDR data was used to identify areas where there was a higher risk of transmission of Zika virus during the 2016 outbreak.

CDR data can also play a role in estimating the current incidence of disease (the number of cases per capita).

Understanding incidence is especially important in the elimination of infectious diseases such as malaria, which is a high priority in many low- and middle-income countries.

In order to reliably estimate the incidence of an infectious disease, we need sufficiently accurate data on the number of cases in the area and the number of people in the area. The number of people is traditionally estimated using projections based on census data. However, this does not account for fluctuations in the population of an area, for example due to seasonal migration. In Namibia, where populations are highly dynamic, this was shown to result in estimates of incidence differing by up to 30% when using a static rather than dynamic population distribution.

CDR data is valuable in this context as it can provide more up-to-date estimates of the size of the population in an area, and therefore more accurate estimates of the incidence of a disease.