In January of 2025, three tigers and a leopard died at the Wildlife Rescue Centre in Nagpur within a short period of each other. Their deaths were due to an influenza virus that normally infects only birds, the H5N1 subtype of the influenza A virus. The term ‘bird flu’ applies generally to influenza infections in birds. H5N1 is one example, but not the only one.
In general, viruses are specific to species, although there are a few notable exceptions, such as the rabies virus. Viruses capable of infecting species that are very different from the original host must undergo a ‘spillover event’, in which a mutation enables them to survive and adapt to the new host. Such events are relatively rare.
H5N1 does, in fact, infect humans, although not commonly. It is fatal in about 30 per cent of diagnosed cases. Deaths from an H5N1 infection have usually happened in people in close contact with poultry. There have, as yet, been no confirmed reports of direct human-to-human transmission.
The World Health Organization (WHO), less than a decade ago, suggested that public health systems should prepare in advance for an unspecified ‘Disease X’, with the potential to become a pandemic. Disease X would be a viral disease, transmitted between humans through a respiratory route. We would have no prior exposure to its causative agent and thus very limited immunity to it. The disease would have a high fatality rate. Those who contracted it would remain infectious for a reasonable period of time, sufficient to infect others. There would be no antivirals. Vaccines would also be unavailable in the early days of the pandemic. COVID-19 emerged just a few years after the WHO exercise was concluded. It fit many of these assumptions.
But could another Disease X replay our COVID-19 experience? The current strongest candidate for Disease X would be a type of bird flu, termed highly pathogenic avian influenza (HPAI), which could also infect humans. Among HPAI possibilities, H5N1 influenza would be the most likely candidate.
The initial stages of any outbreak of a novel disease are marked by considerable uncertainty. Nothing would be known about its epidemiology: how fast it spreads between people, how fast it kills, and whether any drug might work against it.
Two bottlenecks are important in the case of HPAI in humans. The first, of course, is the ability of the virus to jump species and cause infections in human hosts. The second is the barrier to human-to-human transmission. The potential to cause a pandemic, once the initial spillover event occurs, depends on the ease with which the virus moves between humans.
Models for how diseases spread can suggest potential disease trajectories. From these we can hope to decipher its epidemiology by comparing to the actual numbers of cases in time,. Models that describe individuals and how disease is transmitted between them are the most granular of such disease models. These models, called agent-based or individual-based models, must be simulated on a computer.
About four years ago, my group began to develop BharatSim, an agent-based model meant to describe the Indian population. BharatSim describes a ‘synthetic’ or computer-made population of agents, with properties that represent the properties of the real population at a statistical level. The synthetic population is obtained by combining a variety of surveys and census information using refined machine learning techniques. BharatSim then performs simulations incorporating networks of homes and workplaces (including schools) as well as the interactions between agents as they move across them.
BharatSim can be used to simulate communities (1,000-10,000 individuals), cities (3-12 million people) and even states (50 million individuals). It is available for free download together with a suite of test programs.
In a paper published recently in BMC Infectious Diseases, we described the initial stages of a potential H5N1 pandemic using BharatSim. We considered a poultry farm and a population in its vicinity. Some members of the population were workers on the farm (primary contacts) while others were their family members (secondary contacts). Additional contacts (tertiary and higher) represented further contacts of those secondary contacts, typically in the workplaces of family members. All were described as individual agents.
An initial H5N1 infection in birds, once initiated through a single infected bird spreads rapidly, especially when they are in close contact. As more and more birds are infected, the probability that primary contacts can get infected increases. Once one human is infected, the next bottleneck encountered relates to whether this initial infection might spread more broadly, including to members of their households.
We explored several scenarios in our simulations, varying infectivity. We examined the conditions under which sustained human-to-human transmission might occur. We showed that the distribution of secondary cases provided important information about how infectious the disease was. This can be quantified in terms of what is called a ‘reproductive ratio’, the average number of secondary cases arising from one primary case. This is easily computed from our simulations.
We incorporated and studied interventions. These included (i) the culling of all birds, (ii) the quarantining of primary and secondary contacts once a threshold of cases is crossed, and (iii) a vaccination drive targeting primary and secondary contacts. We found that culling on farms before the epidemic peaks, was crucial to preventing later spread. Reasoning from the numbers of those infected in a household could provide a good early estimate of the basic reproductive ratio. Once the disease escaped the network of secondary contacts, in a situation of community transmission, it turns out to be impossible to quell without more global interventions such as a lockdown. We studied how vaccination increased the threshold infectivity, at the population level, required for the disease to spread.
We could conclude that it was in the very early stages of an outbreak that control measures made the most difference. Once tertiary contacts and beyond are infected, community transmission takes over. After this, cruder public-health measures such as lockdowns, compulsory masking, and large-scale vaccination drives are the only remaining options.
Projecting the trajectory of a disease from initial fragmentary information available at outbreak onset is difficult, but it is here that our methods are most useful. Because BharatSim allows many parallel scenarios to be explored in real-time, it can allow policy interventions to be explored and refined on the fly.
Dealing with bird flu has economic consequences. India is the third largest producer of eggs in the world, next only to China and the USA. The Union government confirmed 41 H5N1 outbreaks in poultry this year till Aug 2025. This primarily affected 10 Indian states. Of these states, Andhra Pradesh, Tamil Nadu, Telangana and Karnataka lead the country in egg production, while Maharashtra is an important source of poultry meat.
Culling poultry at scale in the wake of a bird flu infection can have major economic costs. Thus, having a clear understanding of what might be at stake is important, since economic needs must be balanced with the dangers of potential pandemic influenza spread. High-quality surveillance, in particular, is necessary to identify birdflu outbreaks. It is also important to track the course of infection in infected humans so that disease severity can be better understood at an early stage.Three of every four emerging diseases come to us from animals. COVID-19 is a prominent example, where the original host is believed to have been a bat. Such diseases are called ‘zoonotic diseases’. Their very existence reminds us that human health is linked inextricably to its ecological and environmental background. The term OneHealth describes this broader context. Simulating disease spread using BharatSim can help capture the complexities of OneHealth approaches, while also revealing what might lie in India’s future.
Gautam I. Menon is a Professor at Ashoka University.
Views expressed are personal and do not reflect the views of the University
Views expressed are the author’s own and don’t necessarily reflect those of Down To Earth