SIAM News Blog

Mathematically Understanding the MERS Epidemic

By Karthika Swamy Cohen

Displaying a map of infectious diseases, Vittoria Colizza questioned, “Why have there been so many epidemics in recent years?”

Increase in population density around the globe is one reason, she went on to explain. With people living closer together, chances of infection are higher. Improved mobility is another. With people traveling farther and wider in shorter time, infectious diseases tend to spreader faster and farther as well. Another reason is the changing environment, which is allowing pathogens to fit in and grow well in new regions. Aging is also a factor –as populations age, more people become vulnerable to infectious diseases.

Colizza, a senior research scientist at Inserm in Paris, focuses on the modeling of emerging infectious disease spread by integrating complex system methods with computational sciences approaches, statistical physics, and geographic information systems. In her invited talk at the SIAM Annual Meeting, “Anatomy of MERS Epidemic,” she discussed the application of these strategies to the study of the Middle East respiratory syndrome coronavirus (MERS), which spread to 26 countries and had 1,638 confirmed cases as of this March, including cases of human-to-human transmission. Mortality from the disease is 35 percent and 75% of cases are localized in Saudi Arabia. 

The virus, which is associated with respiratory tract infection in humans, is a major public health concern since disease spread has been ongoing since the first case in June 2012. 

Colizza’s group conducted work during successive phases of the epidemic to study its transmissibility, including zoonotic vs. human routes, seasonal variability, and geographical heterogeneity. 

Colizza explained the peculiar ecology of the disease. Camels are a likely major reservoir host for MERS and a source of MERS infection in humans. The role of camels in transmission of the virus is not well understood, but birthing, racing, milking, and contact with camels can transmit the disease.

Sporadic infections are acquired from non-human exposures. The virus does not appear to pass easily from person to person unless there is long intense contact.

MERS exhibits generational cases, with the disease dying out within three to five disease generations. This can be seen from the large peaks in the graph. The disease does not impact different regions in the same way. 

In order to understand the disease, it’s important to answer the questions:
--How much is disease spread due to sporadic cases? 
--How much is it due to different types of transmission? 

Imported cases in Europe were human-to-human transmissions. This was true of the first case in the UK, which was brought by an individual returning to his family in the UK from the Middle East. In France, another isolated case of human-human transmission was reported. Colizza started working on the disease at this time to determine if human to human transmission could lead to a large scale epidemic.cases continued to occur and the disease spread. Summer 2015 brought the largest case in a country other than the Middle East - South Korea – where a few hundred cases were reported.

Colizza’s group used an integrative maximum likelihood analysis of cluster data in the Middle East as well as importations in Europe to analyze the transmission situation and incidence of sporadic infections. Their strategy was based on a spatial-transmission model, which integrates mobility data worldwide and allows for variations in environmental transmission and underascertainment. Underascertainment takes into account the possibility that there are additional cases to those not being tracked by surveillance systems, which is always a challenge in emerging diseases. 

The group’s approach involved integrating incidence—number of cases in time—in a global model. The integrative model is based on a combined maximum likelihood analysis to estimate the reproductive number jointly with the daily rate of sporadic introduction of the virus in the population through zoonotic or environmental means.

According to the reported figures the breakdown of cases in terms of sporadic and human to human transmission were two thirds and one third respectively. However, one possibility to consider was that some of the cases assigned as sporadic could have been human to human transmissions. Surveillance may be affected by underestimation in emerging areas.

In order to estimate these errors, the model backtracked in time to determine how many cases from the source country would have resulted in the current number of cases to estimate this number. 

The group extended a method previously used for the estimation of the seasonal transmission potential of H1N1 pandemic based on the global epidemic and mobility model (GLEAM) by factoring in the different transmission scenarios seen in the case of MERS. GLEAM was specifically developed to study human human transmission.

GLEAM integrates different layers of data and models population data with NASA satellite inventory. It also accounts for flights across two areas to incorporate passengers’mobility. Once all these parameters were integrated in a computational model, the group simulated the spread of disease.

Once the number of cases reported in an area is estimated, the group can estimate that as a percentage of total cases. With both spatial and temporal data, they work with a tool of 32 models that were fit to the data.

Using GLEAM, the group was able to estimate the ratio between sporadic and human transmissions. While the WHO had estimated this as 66% (sporadic) to 34% (human to human), the model found the number to be 25% (sporadic) to 75% (human to human). Hence, human to human transmission was found to be much higher than previously thought. 

One of the reasons for this could be that it is harder to identify the disease in subclinical forms. Asymptomatic patients were possibly not identified as MERS patients. This is an additional burden on surveillance systems. MERS symptoms are common across a range of diseases, making it harder to identify and isolate the disease to diminish risk of transmission within hospitals. This indicates that management of cases needs to be well thought out.

Air travel is easily one of the predominant reasons for global spread of infectious disease epidemics, as previously seen with SARS and the 2009 H1N1. The Middle East has a central role in connecting different regions of the world today, having shown a notable increase in traffic growth in the past several years. Aside from seasonal variations due to mass gatherings or entry and exit of expats for jobs, an emerging pandemic in the Middle East constitutes high risk for rapid and worldwide disease spread.

Karthika Swamy Cohen is the managing editor of SIAM News.

blog comments powered by Disqus