Each year as monsoon-bearing clouds arrive over eastern Uttar Pradesh, so does a pestilence. Hundreds, mostly children, fall ill with fever, convulsions, and disorientation. So familiar are people with these symptoms that at the first signs they rush their children to Baba Raghav Das Medical College Hospital, Gorakhpur. Few other hospitals in the eastern Uttar Pradesh region either have the doctors or the equipment to handle this affliction. And so, the journey of children from nearly 15 districts around Gorakhpur ends at the biggest hospital in these parts.

For the hospital, it is an annual, inevitable, slow descent into chaos.

By September, around 50 patients arrive every day, and the medical college’s infrastructure stretches to breaking point. There aren’t enough beds: unconscious children, sometimes two or three, lie on a single bed. There aren’t enough doctors, so diploma students in child health from around Gorakhpur are drafted in. There aren’t enough mechanical ventilators, even though patients with severe encephalitis often need them.

“It is very difficult for us to manage these patients,” says Anita Mehta, one of the six paediatricians at Baba Raghav Das hospital.

The biggest problem here, however, isn’t lack of beds, doctors or ventilators. It is the lack of diagnosis. Nobody at the hospital knows what causes the outbreak year after year.

The illness is called Acute Encephalitis Syndrome (AES), but the term is merely a convenient label for symptoms of an inflammatory brain disease. AES can be caused by any of several viruses, bacteria and fungi. The culprit is never identified in most of the children who come for treatment.

But there are several hypotheses. A favourite among the doctors at Baba Raghav Das, and the Uttar Pradesh government, is enteroviruses, a genus of viruses that includes polio, and which spread through nasal, oral and faecal routes. A handful of studies in the last decade have found evidence of enterovirus infections in Gorakhpur’s AES patients, giving credence to the hypothesis. But other studies have not. Yet many of the state government’s actions to fight AES are geared towards enteroviruses—replacing vulnerable hand pumps to prevent drinking water contamination from faecal matter, for example. The measures haven’t worked and the AES burden remains heavy.

The biggest problem here, however, isn’t lack of beds, doctors or ventilators. It is the lack of diagnosis. Nobody at the hospital knows what causes the outbreak year after year.

There are two explanations for this failure: That AES isn’t caused by an enterovirus. Or the precautions against enteroviral infections were not properly implemented. Several doctors in Gorakhpur subscribe to the second explanation. Dr. R. N. Singh, who taught at Baba Raghav Das Medical College in the Seventies and began running a campaign for encephalitis eradication after he retired, says the government installed a minuscule 3,000 hand pumps across Gorakhpur, while more than double the number would have been needed to minimise water contamination. “Yeh oonth ke mooh mein jeera bhi nahin hai,” he says, using a Hindi idiom for something so small, that it doesn’t even compare with a cumin seed in a camel’s mouth.

The other explanation for why anti-enterovirus measures aren’t working is that the Gorakhpur illness is caused by scrub typhus, a disease endemic to Uttar Pradesh. Scrub typhus is caused by a bacterium transmitted by a microscopic mite that lives in scrub vegetation. Children walking in bushes without proper clothing are especially at risk, as are woodmen or forest foragers, both common professions in the region.

The evidence for this hypothesis—starting with an August 2014 study by the Indian Council of Medical Research (ICMR) that found scrub typhus antibodies in over 60 per cent of the sick children in Baba Raghav Das—is rapidly growing. But doctors at Baba Raghav Das are sceptical. One reason is that scrub typhus is not known to cause encephalitis outbreaks, triggering a long-drawn fever instead. Encephalitis as a complication occurs in a few untreated sufferers of scrub typhus, rather than among hundreds, as is the case in Gorakhpur, says Komal Kushwaha, a paediatrician who worked at Baba Raghav Das for almost 30 years before retiring in 2015.

“Scrub typhus cannot be the reason for almost all unconscious febrile children coming for treatment. When cases come, they are sporadic and we see 3-4 per cent every year. But they cannot cause an outbreak-like situation,” he argues.

Kushwaha, who is firmly in the enterovirus camp, dismisses the typhus hypothesis as a “gimmick” which the government wants to push because they cannot tackle the deep-rooted causes of enteroviral encephalitis.

The scrub typhus camp has a different explanation, though. Arun Kumar, a virologist at Manipal University, who headed the ICMR study, suggests that the so-called outbreak of scrub typhus encephalitis could merely be due to a bias in the way data is collected. Because no data exists on scrub typhus patients who don’t develop encephalitis, and because the most severe cases of encephalitis in eastern Uttar Pradesh are rushed to Baba Raghav Das, the illness looks like an encephalitis outbreak.

“These patients are only the top of the pyramid,” says Kumar. “The bottom of the pyramid is patients with fever, for whom you don’t have any data. If you plot back each patient’s address into a map, you will see only one encephalitis case from a village.”

As the disagreement on the causes of the outbreak continues, the deaths mount at Baba Raghav Das hospital. Last year, around 500 of the 2,894 people with AES died in Uttar Pradesh. This year, over 500 cases have already been recorded at the hospital. For most of these patients, the cause of the illness remained unknown. “For around 80 per cent of the patients who come here, we can’t give a proper diagnosis,” says Mahima Mittal, head of the department of paediatrics at the Baba Raghav Das Hospital.

The Gorakhpur syndrome remains a mystery.

If there is one disease that has consistently been tagged “mystery illness” across India, it is encephalitis, specifically Acute Encephalitis Syndrome (AES). It is a sudden, rapidly worsening version of encephalitis that brings on high fever, coma, and in many cases, death. There are four major encephalitis pockets in India today: Uttar Pradesh, Assam, West Bengal and Bihar.

In each of these states, the cause of illness is unknown in up to 70 per cent of the patients. Indian journals, especially paediatric periodicals, routinely carry editorials lamenting this mystery illness, issuing calls to arms, and hypothesising about the causes. But the burden has only increased over the years. Between 2008 and 2015, the number of people afflicted with AES across India more than doubled from 3,855 to 8,941. Mortality remained at around 15 per cent throughout. In all states, except Assam, the majority of those who fall ill and die are very young children, because encephalitis wreaks havoc with their underdeveloped immunity.

The commonest of these pathogens is the Japanese Encephalitis virus. But it accounts for less than a quarter of all cases in India. In the rest, the only information we have to work with is a long list of suspects.

One major problem is that AES isn’t a single illness, but a broad label coined by the World Health Organization (WHO) about ten years ago, for a clutch of illnesses difficult to differentiate from each other in the chaotic environment of an outbreak. Around 100 pathogens are known to trigger acute encephalitis, with the list narrowing down to around three dozen in India.

The commonest of these pathogens is the Japanese Encephalitis virus. But it accounts for less than a quarter of all cases in India. In the rest, the only information we have to work with is a long list of suspects. The West Nile virus, dengue virus, Chandipura virus and chikungunya virus have been shown to cause AES. Among bacteria, Streptococcus pneumoniaeOrientia tsutsugamushi, and Haemophilus influenzae, to name a few, can trigger encephalitis. Even fungi like Candida albicans have sometimes been implicated in the disease.

These pathogens cause symptoms that look the same at first blush. They inflame brain tissue, causing the patient to behave abnormally, become forgetful, act confused, get drowsy, etc.—a set of behaviours doctors refer to as “altered mental status”. Meanwhile, fever breaks out, and spikes of activity in the brain result in seizures. This triad of symptoms—fever, altered mental status, and seizures—are together labelled Acute Encephalitis Syndrome.

This symptomatic overlap among so many different illnesses means that even in the best hospitals of the world, the causes of encephalitis are found only half the time. Yet there are distinct patterns everywhere.

Asian countries have traditionally seen a high burden of Japanese encephalitis; Australia has seen a re-emergence of Murray Valley encephalitis in this decade; while some of the leading causes in the United States have been the herpes simplex family and West Nile viruses. Among Asian countries, the endemic regions are India, Nepal and Bangladesh in the subcontinent, and Cambodia, Vietnam and Timor Leste, among others, in Southeast Asia. Several mystery encephalitis pockets exist in these countries too.

India, for its sheer size alone, has one of the largest populations at risk, 375 million, in the encephalitis belt, especially Uttar Pradesh. But India also has other disadvantages: poverty and a crumbling health infrastructure increase both the risk of illness and the probability that it won’t be diagnosed correctly.


Paediatric ICU at B. R. D. College and Hospital, Gorakhpur, one of the hot zones for encephalitis outbreaks.
This and tile photo: Prabhat Singh 

The direct link between poverty and AES can be seen in the demographics of encephalitis patients. The overwhelming majority of patients are poor. All the accompaniments of poverty, from poor hygiene and lack of proper clothing to poor nutrition, increase vulnerability to encephalitis pathogens. And when people do get sick, they only have access to public health facilities, most of which can neither diagnose nor treat the affliction as quickly as it should be treated.

In 2013, following the recommendations of an expert group on encephalitis, Uttar Pradesh set up 104 treatment centres to bolster healthcare facilities. The idea was to offer high-end treatment such as ventilators, and trained doctors, so that patients could be stabilised near their homes instead of having to be rushed to Baba Raghav Das.

Potharaju Nagabhushana Rao, a paediatrician from Hyderabad and lead author of the expert group’s report, had recommended this move because the journey to large hospitals such as Baba Raghav Das was killing more patients than the illness.

With no attempts to feed or oxygenate critically ill patients as they were moved from village to hospital, their sugar levels dropped and the brain was starved of oxygen.

“By the time the patient is taken to hospital, the brain is damaged beyond repair,” says Potharaju. This was the problem treatment centres were meant to fix, by providing small but high-tech hospitals closer to the affected villages.

“When a hut catches fire in a by-lane, what will a fire engine do there?” says Potharaju, “When a problem is in a village, you can’t start a tertiary care hospital 600 kilometres away.”

The centres haven’t worked as planned. Patients from across UP continue to flock to Baba Raghav Das, doctors there say, because they believe that it is the only hospital that can provide care. It is, partly, because not all the encephalitis treatment centres are functional. Some are missing doctors, while others are missing critical drugs, says Kushwaha. As a result, mortality continues to be high while diagnosis remains elusive.

R. N. Singh calculates that Baba Raghav Das hospital has an average of over 200 deaths per bed since the outbreaks began in 1977. “Gorakhpur is the capital of the world for encephalitis. There is nowhere else in the world which would have witnessed more than 200 deaths per bed. It is a very sorry figure,” he says.

About 2,000 kilometres south of Gorakhpur, in Bengaluru, are the verdurous grounds of the National Institute of Mental Health and Neurosciences (NIMHANS). Karnataka sees far fewer cases of encephalitis than Uttar Pradesh or Assam, but it is here that the wheels are being set in motion for an ambitious project to find the causes of AES in the encephalitis belt. It’s a project that aims to use hard evidence, collected over several years and locations, to find the pathogens and alter treatment protocols. It’s going to be a long journey.

Sixty-one-year-old Ravi Vasanthapuram, a neurovirologist at NIMHANS, has been researching AES for over three decades now. With salt and pepper hair and the mien of a professor, he speaks softly most of the time, but breaks into a sardonic chuckle when he talks of the inexplicable neglect AES has encountered from the government. For example, he says, the systems for collecting samples from swine flu patients are far better maintained than for collecting blood and cerebrospinal fluid samples from AES patients, though it is a dangerous illness of the brain.

“You get all the information about someone who sneezes with a fever…But when it comes to AES, the cerebrospinal fluid from patients still comes at room temperature, and it doesn’t come in a screw-top vial. These are the realities. The nose gets more importance than the brain.”

Vasanthapuram believes he might be able to change this. In 2014, he received a $4.8 million grant, spread over three years, from the Centers for Disease Control, the United States’ leading public health institute, to study the causes of AES.

With these funds, he upgraded a network of diagnostic laboratories across the encephalitis belt, so that they could now test for 10 different pathogens in every patient sample, including for scrub typhus. In the first year of the project, nine laboratories from Assam, Uttar Pradesh, West Bengal and Karnataka were roped into the network. Karnataka was chosen as a control location, for comparison with the data collected from high-burden regions.

Testing for encephalitis causing agents in India has so far focused on Japanese encephalitis. This is partly because Japanese encephalitis has traditionally made up a large number of cases. Over the years, though, the epidemiology of AES has changed. The proportion of Japanese encephalitis cases has dropped steeply, though the number of AES cases has remained the same.

Despite this, the National Vector Borne Disease Control Programme continues to test only for Japanese encephalitis. When that is ruled out, the illness is labelled AES. This means any of the three dozen agents that cause encephalitis could be hiding in the basket labelled AES. Vasanthapuram calls AES a “dustbin diagnosis”, a label under which you file every encephalitis case for which you cannot find a cause.

It is this “dustbin” Vasanthapuram’s group will sift through over the next three years. By testing for 10 agents the laboratories will give names to some of the illnesses. These agents include the scrub typhus bacterium, the herpes virus, three bacteria that cause meningitis, and the chikungunya virus. Each is capable of triggering an illness that is either encephalitis or looks deceptively like it.

In October 2015, Vasanthapuram’s group shared the findings in an abstract during the ID Week conference of infectious disease specialists in San Diego, US.

In their first study of 1,253 patients in 2014, they found the pathogen in nearly 40 per cent of the cases. As expected, Japanese encephalitis was the commonest cause of AES in the four states, at 21 per cent. Among the non-JE cases, the biggest cause was an agent suspected in outbreaks such as the one in Gorakhpur, but not taken seriously: scrub typhus. Other significant causes were herpes simplex and the meningitis-causing Streptococcus pneumoniae.

In 2005, the threat of Japanese encephalitis hit home. India’s biggest epidemic raged across Uttar Pradesh, killing 1,344 out of the 5,737 affected children. It forced India to begin importing a vaccine from China and to kick-start a massive vaccination campaign in 2007. 

Meanwhile enteroviruses, a favourite hypothesis in Gorakhpur, turned up only around 0.2 per cent of the time in the four states, including Uttar Pradesh. “We are calling the bluff on enteroviruses,” says Vasanthapuram.

The story of AES in India begins in 1956 in North Arcot district (now Tiruvannamalai and Vellore) of the Madras Presidency. In that year, two separate groups reported cases of Japanese encephalitis in the region. One reported clinical symptoms of Japanese encephalitis among children, while the other found Japanese encephalitis antibodies in six patients. Once the virus was isolated in southern India, it began showing up across the country—an outbreak in West Bengal’s Burdwan and Bankura districts in 1973 killed 300 people. Then, between 1977 and 1979, outbreaks occurred in Tamil Nadu, Andhra Pradesh and Karnataka in the south, and West Bengal, Assam, Bihar and Uttar Pradesh in the north. Since then, Japanese encephalitis has become endemic to many of these states.

In 2005, the threat of Japanese encephalitis hit home. India’s biggest epidemic raged across Uttar Pradesh, killing 1,344 out of the 5,737 affected children. It forced India to begin importing a vaccine from China and to kick-start a massive vaccination campaign in 2007. Within five years, 15 states had been covered, and the Japanese encephalitis vaccine became a part of their universal immunisation programme. This was a point of inflection for encephalitis in India, because it sharply reduced the incidence of Japanese encephalitis without touching the other causes of AES lurking in its shadows.

According to WHO, a patient suffering from fever, altered mental status with or without seizures should first be tested for Japanese encephalitis. If the results are negative, it would be an AES case and tested for similar illnesses such as bacterial meningitis and herpes simplex encephalitis. In essence, AES was defined as any illness that looked like Japanese encephalitis but couldn’t be confirmed as such.

India adopted WHO guidelines, but went after Japanese encephalitis more than other causes of AES. Funded by the National Vector Borne Disease Control Programme, hospitals in endemic areas received Japanese encephalitis test kits. But once the results turned out negative, most hospitals could not test for other bacteria and viruses recommended by WHO, and recorded them as AES of unknown cause.

The result of this focus is that the incidence of Japanese encephalitis has fallen while AES from an unknown cause has remained the same.

What are some of the illnesses hiding in the “AES of unknown cause” diagnosis? Sporadic studies from outbreak regions throw up some clues. A study from the Regional Medical Research Centre in Dibrugarh, Assam, for example, found antibodies to the bacterium leptospira in 8 of 197 patients of AES. Leptospira are spiral wormlike bacteria transmitted to humans from infected rodent urine. While the illness mostly manifests as a fever and body ache, the 2012 study showed that it also causes altered mental status, thus being labelled AES. The West Nile virus, first isolated in West Nile district of Uganda in the 1930s, has shown up in Assam. A 2003 epidemic in Andhra Pradesh was attributed to a newly emerging virus named the Chandipura virus. In other outbreaks, researchers also saw the dengue virus, the malaria protozoa and the chikungunya virus causing encephalitis symptoms.

A complete picture of the prevalence of each of these agents doesn’t exist, because the government continues to collect data only for Japanese encephalitis. As a result, “AES of unknown cause” has become the dominant diagnosis in most encephalitis pockets in India.

As encephalitis cases swell in Gorakhpur in July, the outbreak season is fading in Assam. The illness seems to track the path of the monsoon, hitting Assam first, then West Bengal, and finally Uttar Pradesh and Bihar. This is to be expected, says Ravi Vasanthapuram. “Globally, most infections occur after the monsoon because there is better transmission. If the pathogen is waterborne, monsoon leads to contamination. Vector density also goes up after monsoon.”

An extremely common vector for encephalitis-causing diseases is the mosquito, which carries the virus for Japanese encephalitis, West Nile, Chikungunya and dengue, among others. Each monsoon marks the start of a new breeding season for mosquitoes.

Assam carries a huge burden of encephalitis: last year, over 2,000 people were affected, while close to 400 died of it. Even though the monsoon arrives only in June, it rains almost through the year in districts such as Jorhat. Located in the flood plain of the Brahmaputra, Jorhat has numerous ponds and lakes dotting its brilliant green landscape. In Jorhat city, the ponds are replaced by open drains clogged with plastic, a perfect breeding ground for mosquitoes. Pigs live by the roadside; piggeries are big business in Assam because pork is a big part of the local diet. But pigs are also amplifying hosts to the Japanese encephalitis virus. The virus multiplies in their bodies and is transmitted to humans by mosquitoes that bite the pigs.

Located in the centre of this bustling, frequently water-logged city is the seven-year old Jorhat Medical College, a spanking new institution compared to Gorakhpur’s 45-year-old Baba Raghav Das Medical College. But like Baba Raghav Das, Jorhat Medical College is also an epicentre for encephalitis treatment in upper Assam, where patients come from all nearby districts, as well as from neighbouring Nagaland.

The outpatient department of Jorhat Medical College is crowded. The heat and extraordinary humidity—around 80 per cent in July—means a miasma of sweat, sickness and disinfectant hangs in the corridors.

In this environment, the air-conditioned paediatric intensive care unit, which houses four encephalitis patients, offers relief. Despite this, a woman is fanning her child who lies rigid and staring on one of the beds in the ICU. The girl, referred to as “bed number 7” by the doctors and interns, has Japanese encephalitis, and is running a high fever. Next to her, on bed number six, lies 12-year-old Madhusmita Phukan, who doesn’t have the benefit of a diagnosis. She has been in Jorhat three days, and samples of her blood and cerebrospinal fluid have been sent to the microbiology department. They will be tested for malaria, Japanese encephalitis and dengue. But the department will not be able to test for herpes simplex, a fairly common cause of AES in upper Assam, because it doesn’t have the kit.

As they await the results, doctors at Jorhat Medical College are treating Phukan for everything that could possibly cause her illness. The antiviral Acylovir is given in case she has herpes simplex, mannitol to control the raised intracranial pressure that kills Japanese encephalitis patients, and antibiotics in case of bacterial encephalitis and to ward off infections from the catheter and naso-gastric tube that invade her body. Through it all Madhusmita lies unconscious, her small body shrouded in a web of medical apparatus.

Such an aggressive, blunderbuss approach works for most patients, says Pranabjit Bishwanath, the head of department of paediatrics at Jorhat Medical College, because viruses such as West Nile and chikungunya do not have specific treatments anyway. So knowing the causative agent doesn’t make a difference in most cases, because the illness is managed according to symptoms alone.

But one illness for which a diagnostic test would really make a difference is herpes simplex, he says. This is because the drug used—Acyclovir—is far too expensive to give to all AES patients, who typically come from extremely poor families. “If we can diagnose herpes early, it will be much better for us,” Bishwanath says. Then the drug can be acquired with the confidence that it will work.

After waiting a day, Madhusmita’s parents decide to move their daughter to a bigger hospital in Guwahati, the capital, in the hope that she can get better treatment as she continues to remain unconscious. She is shifted with no diagnosis other than the mysterious AES, because the results of the Japanese encephalitis test are yet to come. A month later, paediatricians at Jorhat Medical College have lost track of her status and moved on to new patients who continue to arrive as the season progresses.

If things go as Vasanthapuram hopes, every major district hospital in the encephalitis belt, including Jorhat Medical College, will either have a laboratory or quick access to a laboratory that can test for multiple agents, including herpes simplex.

“What I would love to do is to equip the labs, train them, and place a field coordinator in each district at a salary of something like Rs. 20,000, considering the seriousness of AES. I’m hoping the government will take on this established system and run it like the polio system,” he says.

In the second year of his project Vasanthapuram’s team tested 2,262 patients in four states, bringing the total sample of his study to a significant 3,300. The trends were similar in 2015 as in 2014. Scrub typhus constituted the second largest number of cases among non-JE causes, especially in eastern Uttar Pradesh districts around Gorakhpur. Meanwhile, Assam continued to have a high percentage of Japanese encephalitis because vaccination campaigns haven’t been as widespread as in Uttar Pradesh. But the team is still stuck at a 40 per cent rate of diagnosis.

To get around this, Vasanthapuram will turn to technology. Next year his team will begin using a method called next-generation sequencing to identify the bugs causing encephalitis. Nucleic acids, such as DNA and RNA, in all patient samples will be sequenced and matched with the genomes of known viruses. This technology is more sensitive than the methods used by Vasanthapuram’s group so far, and is expected to pick up the causes in another 20 per cent of encephalitis cases.

One of the things that could happen, Vasanthapuram says, is that more Japanese encephalitis, dengue and scrub typhus cases will show up, simply because the current diagnostic methods are not sensitive enough to identify them. I ask if he expects to find some unknown virus, that hasn’t been known to cause encephalitis in India. “I don’t know,” he says, “Next Generation Sequencing is something very new, and the whole world is experimenting with it. But I believe that from the existing 40 per cent diagnostic rate, we may move another 10 or 20 per cent.” The remainder, he accepts, will always be a challenge—a challenge even the best healthcare systems in the world haven’t overcome.

But some experts ask if high-tech diagnostic tests alone are enough to uncover the aetiologies in mystery outbreaks, if there aren’t enough expert clinicians—doctors on the ground who observe the symptoms of patients—to provide early clues on what to test for.

One such expert is T. Jacob John, an infectious diseases specialist credited with uncovering the aetiology in at least two AES pockets in India, Bihar’s Muzaffarpur and Uttar Pradesh’s Saharanpur. Both investigations highlight how critical a clinician’s perspective is to an AES investigation.

Muzaffarpur is a district in Bihar with a two-decade-long history of recurring AES outbreaks, like neighbouring Uttar Pradesh’s Gorakhpur. The dog days of summer see hundreds of young children fall ill in this town suddenly and seriously. The illness hits in the wee hours after the child has been asleep for some time. Parents often wake up to a sharp cry from their child, to discover the child is convulsing and losing consciousness. They rush the child to the hospital, but the illness progresses to a coma. Within 48 hours, over half of the children are dead, an extraordinarily high fatality rate even for AES.

Since the outbreak was first reported in 1995, it was thought to be and treated as Japanese encephalitis, a practice that did not change even when clues began to emerge that it was neither Japanese encephalitis, nor even encephalitis. In 2013, frustrated by the large number of deaths each year, Muzaffarpur-based paediatrician Arun Shah invited John to help investigate the illness.

I was horrified to discover that none of the children showed any evidence of encephalitis.

John, who refers to himself as a freelance medical investigator since he retired from the Christian Medical College in Vellore in 1995, agreed to do so. When he arrived in Muzaffarpur in May 2013, he began exploring the town, interviewing parents of affected children, taking in the surroundings. Next, as the season began, he scoured the hospital records. Something significant struck him.

“I was horrified to discover that none of the children showed any evidence of encephalitis,” he says.


There is such a shortage of beds at the Gorakhpur hospital that sometimes children to a three share a bed during peak season.
Photo: Prabhat Singh.

Over the years, John has published several articles emphasising the need to differentiate encephalitis from encephalopathy. Both illnesses look alike and affect the mental status of patients, but careful clinicians can distinguish between them through symptoms and basic laboratory investigations. In encephalitis, which is primarily an illness of the brain, for example, the patient shows focal neurological signs, or evidence that a particular region of the brain is affected. This manifests in symptoms such as an unsteady gait, which points to an impaired frontal lobe. In encephalopathy, where the brain is a secondary victim of the illness, patients often don’t show these symptoms.

Despite the differences between encephalitis and encephalopathy, researchers have, time and again, conflated both during epidemics, to the detriment of diagnosis. Knowing whether an illness is an encephalopathy or encephalitis can be a valuable first clue in finding the cause, because one can trace it to a virus or a bacterium known to produce this symptom. For example, if it is an encephalopathy, doctors can usually rule out Japanese encephalitis and West Nile, because these viruses are neurotropic, meaning they hit the brain first.

The confusion between encephalopathy and encephalitis has a long history in India. An outbreak in Nagpur during the 1950s was thought to be encephalitis until a team from ICMR, and another of which John was part of, clarified that it was an encephalopathy. Then in the early 2000s John, along with a team of doctors including paediatrician Vipin M Vashishtha from Bijnor district of Uttar Pradesh, set about investigating recurring encephalitis outbreaks in Saharanpur, a district about 100 kilometres from Bijnor. This Saharanpur disease, unlike other AES epidemics across India that occur during monsoon, was a winter illness, with fatality as high as 70 per cent.

Again, John and Vashishtha found after an examination of patients that the disease was an encephalopathy, even though previous investigators, including the ICMR, had labelled it encephalitis. But John and Vashishtha found signs that ruled out encephalitis: the children did not have raised levels of cells in cerebrospinal fluid, there was no evidence of inflammation in the brain, and they tended to contract the illness suddenly, with no warning. Encephalitis patients usually experience prodromal symptoms, or early symptoms of malaise, before the illness worsens.

During an investigation that lasted between 2003 and 2005, both researchers also examined tissue samples from 55 of the children in Saharanpur. They found that the encephalopathy was triggered by damage to the liver and muscle tissue, probably due to a toxin. Instances of liver damage due to the consumption of certain toxic plants had previously been reported from various parts of the world. For example, a European mushroom species, commonly referred to as Death Cap, was known to trigger the kind of liver damage John and Vashishtha observed in their patients.

Next, John’s team interviewed families of the sick children, to find that these children sometimes ate a bean growing on a local plant called Cassia occidentalis and known in Hindi as pamaad. Analysis of the plant revealed a potent toxin which could cause the liver damage and encephalopathy found in the children.

When they shared their findings with the Saharanpur administration, it organised large scale awareness campaigns and drives to uproot the plant. Over the years, the illness disappeared from Saharanpur.

Having been through this experience, the Muzaffarpur illness was déjà vu to John. Once again, the illness was an encephalopathy, although the main co-symptom here was hypoglycaemia or low blood sugar, rather than liver damage. John noted that previous studies, which had ruled out Japanese encephalitis and other viruses as possible causes, had found the illness was strongly associated with the lychee season. A similar illness occurred around the same time in northern Vietnam and Bangladesh, both lychee-producing countries. So, researchers there were investigating viruses in bat excrement that sometimes fell on lychees, as well as pesticides used on lychee fruits, as possible triggers for the illness.

As far as John could see, however, all symptoms were pointing to a plant toxin. His own research into plant toxins during the Saharanpur case had introduced him to a Jamaican fruit called Ackee which, consumed in large amounts, caused an illness called the Jamaican Vomiting Sickness. Now, he recalled that the Muzaffarpur encephalopathy looked a lot like Jamaican Vomiting Sickness, down to the hypoglycaemia and encephalopathy. The question now was: were Muzaffarpur and Jamaica connected in some way? A study of published journal papers threw up the answer.

“Do you know what I found just through an armchair literature survey?” John says with a chuckle. “Ackee and lychee belong to the same family. Nobody had looked for this association earlier.”

Like the seeds of the Ackee fruit, lychee seeds also contained a toxin which had been shown to kill rats by triggering encephalopathy. So John began testing if the treatment for Jamaican Vomiting Sickness—an injection of the sugar dextrose—would work in the Muzaffarpur patients. It did, as he reported in a correspondence published in 2014 in the Indian journal Current Science.

A year later, John’s hypothesis found further support. The Centers for Disease Control, collaborating on an investigation in Muzaffarpur with Delhi’s National Centre for Disease Control, also implicated the lychee toxin after two consecutive studies in 2013 and 2014. John has pursued the hypothesis further and tested the flesh of lychee fruits for the toxin, because it was unclear how children could ingest the toxin if it was only present in the seeds. As he expected, the translucent flesh of the lychee fruit did contain traces of the toxin.

Both the Muzaffarpur and Saharanpur investigations show the importance of studying patient symptoms, or framing a “case-definition”, rather than merely relying on lab diagnostic tests, says John. “The problem with looking for aetiology without a case (definition) is the same problem that was seen in Muzaffarpur and Saharanpur. For 20 years, ICMR, NIV and NIDC (now called National Centre for Disease Control) kept coming, collecting samples and looking for causes in them.” Each of these institutions assumed the illness was viral encephalitis, and kept looking for the purported virus. But a quick look at symptoms would have revealed that illness was neither encephalitis, nor viral.

John and several other researchers, including Vashishtha, have been calling for a change in the definition of Acute Encephalitis Syndrome. AES is too broad a label, they say, because any illness meeting the criteria of fever, altered mental status and seizures is classified as AES. This means many encephalopathies caused by non-viral agents are also classified as AES by inexperienced doctors who don’t focus on case-definitions. This is harmful, John argues.

The term Acute Encephalitis Syndrome was never meant to be a diagnosis. It was merely a surveillance tool created by WHO so that it would not fail to count cases that could be Japanese encephalitis, but couldn’t be confirmed because of a lack of access to lab testing kits.

“In other words, AES means that Japanese encephalitis is most likely. It is purely a surveillance terminology. It has crept into clinical diagnosis, unfortunately, and now people are using it as a diagnostic category,” he says.

When I ask John about the scrub typhus controversy in Gorakhpur, he says he finds it shocking that scrub typhus would be confused with AES. Again, he suspects, the doctors who studied the patients before Kumar’s study did not focus on a case definition. If they had, they would have pieced together the symptomatic history of patients, and discovered that they had suffered a long drawn fever, had characteristic rash, liver and spleen enlargement, and other symptoms of scrub typhus.

“If you knew the antecedents of that problem: fever and rash, you wouldn’t call it Acute Encephalitis Syndrome. You would call it scrub typhus, a disease known all over India. It tells me that the whole thing is a wild goose chase, and nobody knows what the goose looks like,” says John.

Vasanthapuram agrees that good clinicians are key to any disease investigation. However, he argues, such exceptional clinicians cannot be present everywhere. “The best way (to diagnose) is to collect specimens. If it is negative for Japanese encephalitis, put it to other tests. The ultimate evidence is a laboratory confirmation,” he says.

Nevertheless, his project isn’t neglecting case definitions. Starting in 2016, his team will collect granular data on the symptoms of patients, along with lab diagnosis, at two of the sites in his network—Dibrugarh in Assam and Deoriya in Uttar Pradesh. Over time, this surveillance will be extended to other sites as well.

Currently, the symptom data collected from all AES patients in India is entered by doctors into a one-page form designed by the National Vector Borne Disease Control Program. The form has a binary format for most questions. Doctors must record whether fever is present, whether vomiting is present, whether seizures are present, whether the patient is vaccinated against Japanese encephalitis, etc. If fever is present, details such as the severity and length of the fever are not noted, even though this data is crucial to a case-definition.

“The form cannot tell you much. It is basically a ticking form,” says Vasanthapuram. His team’s surveillance in Deoriya and Dibrugarh will capture much more: the type of fever, length of fever, the results of other biochemistry tests and the results of computed tomography (CT) scans, for example. This data could some day supply the missing case-definitions, at least for some regions in India.

“This is hard data,” he says, “and people will be convinced only when they see hard data. Without hard data, there will only be stories going around.”

In Gorakhpur, the confusion over the causes of AES continues.

On the one hand, evidence for the scrub typhus hypothesis is growing. After the ICMR study in 2014, Vasanthapuram’s group also reported evidence of scrub typhus in districts surrounding Gorakhpur from their studies. John, who visited Gorakhpur recently with ICMR, believes the hypothesis to be plausible too.

 On the other hand, the accepted treatment for scrub typhus—intravenous azithromycin—doesn’t seem to be working for the patients in Baba Raghav Das Hospital.

But the ICMR believes this doesn’t disprove the hypothesis. One reason that azithromycin isn’t triggering a dramatic recovery in AES patients could be that these patients come to the hospital too late to respond to treatment, says a scientist who does not wish to be named from the ICMR’s National Institute of Epidemiology. Another could be that Baba Raghav Das hospital hasn’t systematically studied the effect of azithromycin in patients, and is missing the improvement in patients who do respond. In either case, more evidence stacks up in favour of the hypothesis than against it, he says.

This is why ICMR continues to bet on scrub typhus. The body has already recommended that Uttar Pradesh’s primary health care centres store doxycycline, an inexpensive antibiotic that can cure scrub typhus quickly, just like azithromycin. The idea is that encephalitis of the severity seen at Baba Raghav Das hospital may be too intractable for azithromycin or doxycycline. Instead, patients must be treated early, while they have nothing more than a fever, so that the illness doesn’t progress to encephalitis. But in the absence of a consensus on scrub typhus, the UP government hasn’t implemented the recommendation yet.

Mahima Mittal, the head of the department of paediatrics at Gorakhpur, doesn’t know what to believe right now. “It is for the researchers to comment on the cause. They give us a diagnosis and we believe it. Maybe now, they will give us a different diagnosis,” she says.

For now, she will treat her patients for symptoms, as she has been doing for years.

(The cover story of the September 2016 issue of  Fountain Ink.)