Our car rumbles beside glistening rice paddies and through Vietnamese villages one house deep. We pass a duck pond, chicken soup shacks, egg sellers, goose-down plumpers, makeshift poultry slaughterhouses snicking and chopping. Improvised electric plucking machines thrum, their stiffened rubber fingers snatching the feathers from broken-necked fowl. Unwanted butchers' ooze mingles with rainwater and sluices along a communal sewer.
Drums and cymbals crash all around as children crowned with red rooster feathers press their faces to the taxi windows. Tomorrow is the 15th day of the eighth lunar month, Tet Trung Thu, the mid-autumn festival when farmers celebrate gathering in the harvest with sweet moon cakes and fables. This year is especially auspicious because Tet Trung Thu falls in the Year of the Rooster, an event in the calendar that resonates in a country founded on poultry.
The chicken is king in Vietnam, and the casual visitor to villages such as this, in the northern province of Thai Binh, in the flood plains of the Red River, east of Hanoi, would never guess that they are on the frontline of a titanic microbiological war. At least three-quarters of all ducks and a quarter of chickens here are now host to a mutable and deadly form of avian influenza which Dr David Nabarro, the UN's newly appointed coordinator for avian and human influenza, warned in September was capable of killing up to 150 million people in a pandemic that "could happen at any time".
Nabarro was accused of alarmism, but his fear is rooted in critical new evidence emerging from the north of Vietnam which has become a breeding ground for the bird flu virus, thanks to fenceless, free-range rearing, communal live markets and back yard slaughterhouses which place people close to the flock and expose birds to anything that flies by.
Since it was first detected here in December 2003, H5N1, to use the scientific shorthand, has quickly spread to South Korea (in December 2003), Japan, China and Thailand (January 2004), Indonesia (July 2004), Cambodia (February 2005) and North Korea (March 2005). By July this year, the virus was in Laos and the Philippines, and spanned the whole of Asia.
Earlier this year, it spilled towards Europe, spreading along migratory superhighways after infecting 184 species of wild birds that nest in the remote lakeland of Qinghai, in north-western China. Russia and Kazakhstan reported bird flu in late July; Mongolia and Tibet in early August. That month, following fears that the virus had crossed the Urals, the Dutch and German governments ordered all poultry to be brought indoors, while ornithologists advised the British government that infected wild birds might land in Norfolk. H5N1 shot up the UK risk assessment table to the highest priority.
However, for all of these apocalyptic stories, as long as bird flu remains in poultry in its current form, it represents a threat primarily to those who farm or slaughter fowl. The critical phase that presages a pandemic is when the virus mutates, becoming readily transmissible between humans. This is what experts think happened 87 years ago when a similar strain of bird flu - misleadingly known as Spanish influenza - engulfed the world with catastrophic consequences. Tissue samples from the time, rediscovered in the Royal London hospital by John Oxford, professor of virology at the London School of Medicine, showed the 1918 pandemic was probably caused by avian influenza that leapt from poultry kept in British army camps on the western front in winter 1916. Within two years, the virus had mutated, enabling it to pass between soldiers as easily as the common cold, killing more than 40 million people.
The current bout of bird flu in northern Vietnam has begun to mutate in a similar way, and the critical phase may have started. At least eight human clusters have been discovered there this year; a number of the victims had no contact with diseased poultry and are likely to have been infected by sick siblings, friends or in hospital. As this human-friendly strain becomes more efficient, it will spread around the world as fast as the strain now endemic in waterfowl.
The first sign that the virus strain had mutated came when international investigators discovered it had leapt the species barrier. Most viruses are unable to adapt to the cellular structure of species other than the one in which they first evolve. However, bird flu quickly learned to reconfigure itself.
First, in August 2004 in China, it hit pigs, whose immune system is closely related to humans'. It moved on to tigers and leopards (October 2004), killing 23 in a central Thailand zoo, then, in Vietnam (June 2005), to domestic cats and rare civets, which in 2003 spread the human outbreak of Severe Acute Respiratory Syndrome (Sars), another virus that originates in poultry. "Every species leap [by H5N1] represents a new virus mutation, increasing the chance that one will become highly infectious to humans," says Dr Peter Horby, an epidemiologist at the World Health Organisation (WHO), in Hanoi.
In northern Vietnam, H5N1 first jumped to humans in January 2004. The virus triggered an explosive cytokine storm in the body as the immune system was induced to attack human tissue, causing massive haemorrhaging in the lungs. The x-rays of one patient in its grip are chilling: on day five, the bottom left of her lung is partially fogged; by day seven, both lungs have filled, as she drowned in her own body fluids.
Within 18 months, 87 people in Vietnam had been infected; 38 of them died excruciating deaths. With each new case, the virus advanced, infecting larger clusters of patients, remaining for weeks in one community before re-emerging in another without visible links. In other Asian countries, including Thailand and Hong Kong, outbreaks among humans were quickly brought under control, but in Vietnam the virus lingered, its ever-changing symptoms making it harder to diagnose and treat. "We are at a critical situation in Vietnam," Horby says. "There are little alarm bells ringing everywhere in Vietnam. The pandemic is inevitable - and probably soon."
However, rather then allow epidemiologists free access to its H5N1 hotspots, the government in Hanoi balked at sharing biological samples with foreign scientists and instead went on a propaganda offensive. This autumn, it announced an £18m poultry vaccination programme which it claimed would protect all birds from the virus. It said it had recruited 10,000 animal health workers to scour the countryside for diseased birds. The daily newspaper Labour told readers that no new cases of human infection had been registered. State radio announced that a new treatment regime, using oseltamivir, an anti-viral drug currently being stockpiled across Europe and the US, had enabled doctors to "cure" dozens of victims. According to one recent government statement, Vietnam had cut the bird flu death rate in humans from 70% in 2004 to 20% this year. Last month, government scientists revealed they were on the verge of going into clinical trials with a radical human vaccine. The country was working hard and winning the war without foreign intervention, Hanoi said. The threat of a pandemic had subsided.
Our car lollops along beside the festival revellers and scores of chickens crammed into coops ready for market as Vietnamese consumers return to eating poultry after two years of fear. We have with us a farmer we picked up 60 miles back at Counter A145 of Hanoi's Hang Da market, where he had just sold a cage-full of live birds. He agreed to show us his farm and now motions for us to take a right turn. The taxi launches itself over the wall of a dyke and skims across the paddy. We abandon it in the mud and wade into a large pond of reddish water filled with geese and ducks. "More than 600, I think," the farmer says, throwing fistfuls of seed into the air.
He pulls a plump duck out of the water, inspecting it carefully and signalling that it is a perfect specimen, unaware that many of these birds are likely to be H5N1 carriers, shedding the virus into the water they paddle in and on the ground they defecate on, and potentially passing it on to the humans who share their environment. The highest risk of all is at the moment of slaughter, when fine particles of poultry blood, mucus and faeces are sprayed through the air.
The farmer wades with his trophy over to a platform where a fire is fizzing. We realise he is going to slaughter the bird to mark our visit and wonder whether we should have brought protective gloves, hats and masks. The farmer folds back the duck's neck, slits its throat, and blood fountains into a bucket. "Don't worry," he says. "Vietnamese birds are now safe. It's the imported Chinese ones, pumped full of chemicals, that carry disease. Vietnam has won the war. Our doctors have beaten bird flu."
Having seen off the French in 1954 at the siege of Dien Bien Phu and the Americans 21 years later with the fall of Saigon, Vietnam is prickly about foreign intervention. Before visiting, all foreign reporters have to propose an itinerary that is then hotly debated and policed by a ministry of foreign affairs minder. It is a costly endeavour: up to $300 a day for the minder, monetary gifts for provincial party apparatchiks and district People's Committee chairmen, small change given before entering anyone's house.
We were granted a visa in 48 hours, and one day later were sitting in a minibus, heading out into the countryside around Hanoi. We had put together a list of potential bird flu hotspots to visit; almost all were rejected by the ministry as being "too far", "speculative" or "unconfirmed cases". However, the government agreed to accompany us to review a recent human cluster in Thai Binh province, a place that has endured more H5N1 outbreaks than any other. "We want you to see we are battling and rapidly winning against bird flu," our minder said.
We had left the capital at dawn, but it was mid-afternoon before we arrived at our destination, having collected on the way representatives from the People's Committee (External Affairs Division), the Preventive Medical Centre, the People's Committee (Foreign Affairs Division), the Department of Protocol (Press Division), the Department of Animal Health and the Ministry of Agriculture. In the remote village of Thuy Lung, the Nguyen family were waiting as the light began to fade. Their cinderblock house was ringed by piles of shaved bamboo that they painted and plaited into temple offerings, many thousands of which would make a couple of dollars. They mumbled begrudgingly for us to come inside, where a young man, watchstrap thin, was sitting, his chest concave, his double tracheotomy scar bobbling as he gasped for every breath. He introduced himself to us like a worn-down scientific specimen. "Tuan. Aged 21. Confirmed H5N1 victim, February 27 2005. Quite well now, thank you."
Three members of his family had come down with bird flu that February. Tuan said he had no idea how he got the virus. His father, Nham, interrupted: "It was last lunar new year. We bought two chickens and a goose. Tuan and I slaughtered them there." He pointed to a concrete wall. "We ate them for three days. He became ill and was sick for 10 days before I took him to hospital on the back of a motorbike on February 20. My son had a fever and was coughing."
The local clinic moved Tuan to the district hospital which referred him to the provincial hospital which sent him to the Bach Mai tropical medicine hospital in Hanoi, where all bird flu cases from northern Vietnam are treated. Tuan remained there until May 13.
Who else had become sick? "My daughter became ill on February 21," Nham said of Ngoan, Tuan's 14-year-old sister. "Maybe she got it from the birds, too." Our minder nodded in agreement. "She was rushed to Hanoi, but had a mild infection and was discharged two weeks later. Both my children were cured."
That left one more infection in the family. "Tuan's grandfather was also infected," Nham said. "He was not really ill, no fever, coughing or headaches. My son cannot go out to the rice fields yet." He'd left hospital weighing only five stone. "But he can look after the TV and video machine here." Nham pointed to a dusty Chinese cabinet behind him, then to a photo of Tuan before he fell ill. The full-faced boy dressed in a dark suit and brogues, out with friends in Thai Binh, was barely recognisable. The minder turned to us: "This is the story of how doctors cured one of the most severe cases of bird flu to hit Vietnam."
However, there were crucial aspects of the story that the minder did his best to gloss over. Tuan was certainly infected after helping his father slaughter birds - but his sister only became sick 11 days later and had no contact with the live or slaughtered fowl, other than eating the cooked meat, from which you cannot contract H5N1. The virus can live in the environment for only 48 hours outside a host, so the sister must have been in close proximity to someone carrying the virus, such as her brother. Here was a likely case of human-to-human transmission, a critical pre-pandemic stage.
Tuan's grandfather was found to be H5N1-positive 17 days after the infected birds had been slaughtered, and 15 days after Tuan had become sick. He had no contact with the slaughtered poultry, so was also likely to have contracted the virus from Tuan.
No one volunteered any information about the fourth victim in the Thuy Lung cluster. Scientific reports we had seen mentioned that a nurse had fallen sick with suspected bird flu at Thai Thuy district hospital on February 27, seven days after Tuan was admitted. This case was presented as separate from the Nguyen cluster. We asked to visit the hospital. "It's closed today," the minder snapped. Why would a hospital close on a Saturday? He scrambled for an answer. "Maybe I can let you go on Monday."
Two days later, we arrived at the hospital in a minibus full of officials, scattering a group of women in conical straw hats. Ushered into a vast committee room, we sat below a bust of Ho Chi Minh, long-deceased founder of the Vietnamese communist party, to hear Dr Dao Trong Bich, the hospital deputy director, repeating a familiar refrain: "This was the most severe case of H5N1 in Vietnam. Our doctors struggled hard and cured him after 82 days in hospital." A pale young man entered the room and shook our hands. "Nurse Thinh, 26, was sick but now 100% fit and can work every day," Bich said.
How did Thinh think he had been infected? He shrugged: "I live in the hospital dorm and couldn't have come into contact with poultry, only nurses and patients." With Tuan? "Oh yes. On February 20, I looked after Tuan for seven hours until I was asked to separate him from other patients." Bich grimaced.
Did Thinh follow any special procedures? He looked nervously at Bich. "I followed procedure." What was it? "To wear protective clothes." A long pause. "A white hat ... and a mask ... maybe gloves?"
Bich chipped in: "We are trying our best, but we do not have facilities to treat H5N1. We have a small supply of oseltamivir, 200 pills." Enough for a handful of patients. "We are not frightened. We must follow the guidance given by the provincial centre."
We approached the provincial authorities at Thai Binh's preventive medical centre. The director who had agreed to brief us was called away, leaving his deputy unprepared. A secretary hastily showed vice-director Dr Bui Thu Hien a pile of notes. He read: "The first cases in Vietnam of humans infected with H5N1 were identified in this province on January 17 2004. After the epidemic started, it began scattering and we had several epicentres. Up until March this year, there have been four clusters in this province, involving 10 patients, half of whom died."
The secretary whispered furiously in Hien's ear, as if to hush him up. But he continued: "1) January 2004, here in Thai Binh city, one brother and two sisters died. The brother's wife was also infected but survived. A fifth member of the cluster was identified 10km away. 2) The following December in Kien Xuong district, three brothers were infected, with one showing no symptoms, another becoming sick and a third dying. 3) This February, a cluster in Thuy Lung village with a 21-year-old, his sister, their grandfather and a male nurse at the local clinic."
This was the Nguyen family cluster and Hien had confirmed that all four infections were interlinked. He continued: "The fourth cluster was in February this year in Kien Xuong district, where a 69-year-old and his wife contracted H5N1. He died and his wife survived. He kept chickens in his back yard, but none of them was sick and so none was culled."
What had been concluded in each case? Were there any common denominators - a poultry breeder who supplied all the families? Hien leafed through the notes and stroked his chin. "There are no definite conclusions. But we stay vigilant. This is a threat not only to our community but to the whole world."
To his left, unmentioned, was a wall-mounted map of the province, dotted with the path of the virus: black triangles (outbreaks in birds); purple spots (human deaths, 2005); red triangles (human cases with symptoms); red spots (humans with no symptoms); red triangles with yellow spots (human deaths, 2004). Thai Binh was pockmarked with bird flu.
There were doctors and bureaucrats who had grasped the scale of the crisis, but whose ability to respond was limited by resources, expertise and cultural attitudes. When bird flu first emerged, the National Institute of Hygiene and Epidemiology (NIHE) in Hanoi, the leading analytical centre, had no safety cabinets, freezers, centrifuges or incubators, and had to turn to the WHO for a £30,000 loan. "We did not even have masks and gloves," said virologist Le Thi Quynh Mai. Even today, the Bach Mai tropical medicine hospital in Hanoi, where Tuan and dozens of other bird flu victims were treated, still does not have a proper isolation ward. As problematic was the adversity in Vietnam to postmortems - no autopsies were consented to by the families of the 38 bird flu victims.
When signs emerged that a pandemic was imminent, every country was asked to report all bird flu cases to the WHO. But Vietnamese scientists were reluctant to pass on their data and the WHO rebuked Hanoi for dragging its feet. Researchers in Britain, China, Canada, Japan, Italy and the US complained that hundreds of throat swabs and serum samples from bird flu victims and their relatives were locked away at the NIHE. Guan Yi, a virologist at the University of Hong Kong, who had been tracking the evolution of H5N1 after it briefly emerged there in 1997, told us, "We need samples almost in real time and that is not happening."
The limited data that virologists in Britain, the US and Hong Kong were able to extract from Vietnam further contradicted the claim by Hanoi that it had the virus under control. The much-touted fall in deaths from bird flu in Vietnam, coupled with an increase in mild H5N1 infections and asymptomatic cases, was a sign that bird flu had further mutated and was now entering an increasingly deadly stage.
John Oxford, at the London School of Medicine, explains, "Any virus needs live hosts if it is to be successfully promulgated around the world. To kill soon after infection would bring a rapid end to an epidemic. When H5N1 begins to kill fewer and infect more, as it is now, a pandemic is potentially close by."
We left Thai Binh preventive medical centre and asked to visit the other clusters revealed by Hien. But in every case the minder found a reason why we couldn't go. "In the first cluster, the wife ran away; we don't know where she is." And the parents of the three dead children? "They have gone to live somewhere else." And so on. We sat in silence as the minibus crawled back to Hanoi.
A day or two later we persuaded the reluctant minder to help us track down that first family. In a nest of alleyways behind a chicken noodle shop in Thai Binh city, a friend of the bereaved family, a member of the local People's Committee, led us to the small door of their neighbour's house. We were beckoned inside. An elderly woman poured green tea and explained how the mother had been driven mad by grief. "This was the first terrible bird flu cluster in the province but the authorities panicked and did not know what to do. My neighbour lost her son and her two daughters in no time at all. And last week her mother died, too. Wait, please, and I'll see if anyone can bear to see you."
The old woman reappeared with sticks of incense. "You can honour their loss," she said, pointing to a door on the opposite side of the alley. We knocked and Le Thi Sang let us in. Her neighbour motioned for us to sit on small stools. Sang said she would never get over the loss of her three children; her son went first, then days after he was buried, her two daughters, who had looked after him in hospital, became feverish. "They found it impossible to breathe," Sang said. "Then they were gone." We saw three bicycles still parked in the back yard.
Did anyone come to investigate the deaths? "They came. Told me nothing and asked me nothing. Said they were sorry and that they would keep people from bothering me."
She pointed up to the corner of the darkened room, to where she had built a shrine with three school photographs draped in black velvet ribbons. Beneath were their dates: Hong, 1974 to January 2004; Hung, 1973 to January 2004; Hanh, 1981 to January 2004. Our incense sticks crackled and glowed as they were lit and added to the offerings - pomelos, pomegranates, a bottle of Vietnamese champagne - peppery smoke coiling around the story of her loss.
The WHO began to suspect it was not getting the whole story from Hanoi last spring and took emergency measures, mounting an inspection tour of the clusters of bird flu cases from northern Vietnam. What the team found was alarming: new bird flu strains had emerged. In several cases, people were carriers without becoming sick. In others, patients were infected but developed symptoms never before associated with the virus - such as diarrhoea and vomiting - leading to misdiagnosis and resulting in an unknown number of victims being sent back out into the community while still highly infectious.
In May, an emergency meeting was convened at the WHO's regional office in Manila, in the Philippines. It concluded: "The pattern of disease appears to have changed [in north Vietnam] in a manner consistent with the possibility that human-to-human transmission has occurred." If action is delayed until there is unmistakable evidence, it most likely will be too late to prevent large numbers of infections and deaths.
By September the WHO was troubled by news that the country's poultry vaccination programme had failed. Last year Vietnam slaughtered more than 55 million birds after H5N1 first emerged, at a loss of $190m or 0.5% of the country's gross domestic product. But since there were still 220 million more fowl in the country, in September Hanoi invested heavily in the inoculations.
Dr Hoang Van Nam, deputy director of the department of animal health, told us his officers had begun with two pilot projects that had both now failed. "The trial programmes were supposed to have been completed before the lunar new year when people buy more chickens and ducks than at any other time. But we have a long way to go." In one pilot area, poultry in only 38 out of 127 communes had been vaccinated. In the second, 1 million fowl were still to be injected. Nam said in both pilots his officers found it hard to distinguish between birds that had been vaccinated and those that had slipped through the programme, making it impossible to work out which were to receive a booster 28 days later, without which the vaccination would fail. No one was supposed to eat or sell birds that had been vaccinated for 30 days after the first injection, but without records or tagging, this rule was unenforceable.
In September, the WHO circulated a report to governments around the world: "Events unfolding in parts of Asia... have sounded a general warning that a pandemic might be imminent. All countries should undertake urgent action to prepare ... to reduce morbidity, mortality and social disruption."
What does a government do when faced with such an immeasurable threat? In the past few weeks, countries have been advised by the WHO to expand morgue space and intensive care facilities, to be ready to close schools and prepare strict quarantine measures. Mass events such as football matches would have to be suspended in the event of a pandemic. In Britain, GPs received 50-page pamphlets from Sir Liam Donaldson, the chief medical officer, with advice on how to spot the symptoms of bird flu that department of health models demonstrated could easily kill up to 650,000 people in the UK alone.
Professor Michael Osterholm, a director at the US department of homeland security, told us, "What people have to understand is that H5N1 will change the world overnight. Foreign trade and travel would be reduced or ended. Transportation would be curtailed. Global, regional and national economies would come to an abrupt halt. I have been accused of being too shrill, scaring people out of their wits, but what I am trying to do is scare them into their wits."
Many governments have begun work on emergency pandemic plans, and one of the key issues they are dealing with is the availability of antivirals. Oseltamivir is in incredibly short supply, with sufficient doses for only 40 million people worldwide - one quarter of the British population being covered, but less than 1% of that in the US.
In any case, oseltamivir is highly problematic. The drug sold under the name Tamiflu has proved hard to administer, requiring large doses, and is ineffective unless prescribed soon after infection takes hold. One recent simulation of the pandemic conducted by Imperial College, London, showed the antiviral would have to reach 80% of the affected population "within three weeks" of an outbreak - something of an impossibility given the drug's short supply and the speed at which the virus will move.
In Asia, the virus has even begun to circumvent oseltamivir: a patient from north Vietnam is resistant to the drug and Tokyo University has reported that 18% of those treated with it became resistant.
The good news is that humans still remain poor conductors for bird flu. The virus concentrates in the lower respiratory system, making it harder to transmit than ordinary human flu, which invades the upper respiratory tract and is easily transmitted through coughing, sneezing or breathing.
The bad news is that in two months' time, the temperature in Vietnam will begin to drop and colds and flus of the common or garden variety will set in. It is also the time of Tet Nguyen Dan, the lunar new year, when most families come together, travelling the length and breadth of the country, buying a live bird to slaughter at home for a celebratory dinner. All it will take is for one person unknowingly infected with H5N1 by a diseased butchered bird to sit on a bench beside someone with a bad case of the human cold and what could emerge is a recombined bird flu virus with the ability to spread around the world at the speed of a sneeze.
